NeFCA
ETMAAL 2013 BOOK OF ABSTRACTS 24 HOURS OF COMMUNICATION SCIENCES
2
ETMAAL 2013
BOOK OF ABSTRACTS
NeFCA, in association with the department of Media and Communication and Erasmus Research Centre for Media, Communication and Culture ERMeCC, is hosting the Etmaal 2013 Conference on Communication Science at Erasmus University Rotterdam on February 7th and 8th, 2013. During the Etmaal, a wide variety of panels and papers on topics relevant to the current state of communication research will be given. The papers are organized into six sections: Culture & entertainment
Health communication
Journalism
Organizational communication
Persuasive communication
Political communication
In this book, you will find all the abstract sorted in these sections. We wish you a good browsing! The Etmaal Organizing Committee Tonny Krijnen David Novak Mijke Slot
2
3
ETMAAL 2013
BOOK OF ABSTRACTS
CONTENTS CULTURE & ENTERTAINMENT .................................................................................................. 14 Session #1 ...................................................................................................................................................... 14 Adolescents’ instrumental and expressive text messaging behavior and their friendship quality: a perceived affordances-approach - Mariek Vanden Abeele, Alexander Schouten and Marjolijn Antheunis ..................... 14 Development and validation of the Adolescent Measure of Empathy and Sympathy (AMES) - Helen Vossen, Jessica Piotrowski and Patti Valkenburg ........................................................................................................... 15 Understanding the role of perceived peer norms in the relationship between sexual media use and young adolescents’ sexual behaviors - Eline Frison, Laura Vandenbosch and Steven Eggermont .............................. 17 Teen sexting across Europe: a cross-national comparison of the predictors of sexting - Susanne E. Baumgartner, Sindy Sumter, Jochen Peter and Patti M. Valkenburg ............................................................... 18 Session #2 ...................................................................................................................................................... 19 Practice makes perfect: the longitudinal effect of adolescents’ instant messaging on their ability to initiate offline friendships - Maria Koutamanis, Helen G.M. Vossen, Jochen Peter and Patti M. Valkenburg ............. 19 How social are social network sites? A longitudinal study among early adolescents - Marjolijn L. Antheunis, Alexander P. Schouten and Emiel Krahmer ...................................................................................................... 20 Linked in or left out - impression formation based on one’s professional and private social network site profile - Christina Jaschinski, Somaya Ben Allouch and Ard Heuvelman .......................................................... 22 Session #3 ...................................................................................................................................................... 24 Sexualizing sitcoms, self-objectification and the sexual double standard: a longitudinal study among adolescents - Laura Vandenbosch and Steven Eggermont ............................................................................... 24 An experimental investigation into the effect of online self-portrayal on women’s self-objectification - Dian de Vries and Jochen Peter ................................................................................................................................ 25 Women’s rejection of sexually explicit content: the role of hyperfemininity and processing style - Johanna M.F. van Oosten, Jochen Peter and Inge Boot.................................................................................................. 27 Media use and ADHD-related behaviors in children and adolescents: A meta-analysis - Sanne Nikkelen, Patti M. Valkenburg and Mariette Huizinga .............................................................................................................. 28 Session #4 ...................................................................................................................................................... 30 The dark side of ‘reality TV’: questions of ethics in the production of ‘popular factual entertainment’ television - Jelle Mast ....................................................................................................................................... 30 Prime time morality: a longitudinal study of morality in Dutch prime-time fiction - Serena Daalmans, Ellen Hijmans and Fred Wester ................................................................................................................................. 31 Biedt Dexter zijn kijkers moreel vermaak? Een geval van “moral disengagement” - Monique Timmers, Ed Tan, Claire Segijn and Guus Bartholome .................................................................................................................. 33 The relationship between fiction viewing and support for the police: the importance of procedural fairness beliefs - Astrid Dirikx and Jan Van Den Bulck .................................................................................................... 34 Session #5 ...................................................................................................................................................... 36 National cultural repertoires of evaluation in a global age: film discourse in France, the Netherlands, the United Kingdom, and the United States - Annemarie Kersten ......................................................................... 36 De (ideologische) representaties van wetenschap in de rampenfilm - Laurens Van der Steen and Pieter Maeseele .......................................................................................................................................................... 36
3
4
ETMAAL 2013
BOOK OF ABSTRACTS
De afbeelding van statusverschillen in sekse en kleur in Nederlandse non-fictie programma’s - Sjoerd Pennekamp, Claire Segijn, Guus Bartholomé and Monique Timmers .............................................................. 37 Film history between culture and economics: admission figures at the Capitole cinema (1953-1971) - Khaël Velders, Liesbeth Van de Vijver and Daniël Biltereyst ...................................................................................... 39 Session #6 ...................................................................................................................................................... 40 A longitudinal study into the effects of adolescents’ social network site use on their appearance investment and desire to undergo cosmetic surgery - Dian de Vries, Jochen Peter, Peter Nikken and Hanneke de Graaf 40 Can the experience of playing a shooter game be as different as chalk and cheese? An investigation of how players construe their playing style - Wannes Ribbens and Steven Malliet ..................................................... 41 Profiel van compulsieve SNS-gebruikers en gamers in België Surveyonderzoek bij volwassenen in Vlaanderen en Wallonië - Rozane De Cock, Jolien Vangeel, Annabelle Klein, Pascal Minotte, Omar Rosas and Gert-Jan Meerkerk .......................................................................................................................................................... 43 Het verband tussen gewelddadige games en desensitisatie: De rol van geobserveerd realisme en speelstijlen - Steven Malliet and Wannes Ribbens .............................................................................................................. 44 Session #7 ...................................................................................................................................................... 46 The relationship between on-line and off-line fear of crime: A mediated model - Kathleen Custers and Jan Van Den Bulck ................................................................................................................................................... 46 Constructing and validating a scale to assess moral judgment of antisocial media content and media preferences among youth - Xanthe S. Plaisier and Elly A. Konijn ..................................................................... 48 The role of media with antisocial content in cyberbullying among adolescents - Anouk H. Den Hamer, Elly A. Konijn and Micha G. Keijer ............................................................................................................................... 49 Double dose: How violent media exposure and family conflict interact to predict adolescents' aggression Karin Fikkers, Jessica Taylor Piotrowski, Helen Vossen and Patti M. Valkenburg ............................................ 51 Session #8 ...................................................................................................................................................... 52 The relationship between parental work demands and children’s television time - Ine Beyens and Steven Eggermont ........................................................................................................................................................ 52 How and why parents guide the media use of baby’s, toddlers, and kindergartners; an exploratory study Peter Nikken ..................................................................................................................................................... 53 Is doorstroom het antwoord? Van basisvaardigheden naar mediawijsheid: drempels en beproevingen binnen een e-inclusie initiatief - Kelia Kaniki Masengo and Ilse Mariën ...................................................................... 55 Mediawijsheid aan de keukentafel. Communicatie binnen Nederlandse gezinnen over media(-wijsheid) Elize Van der Meulen and Jordy Gosselt ........................................................................................................... 56 Session #9 ...................................................................................................................................................... 58 Museums 2.0: A study into expertise and culture within the museum blogosphere - Jessica Verboom and Payal Arora........................................................................................................................................................ 58 Knowledge and innovation in media clusters: the case of Hilversum, Netherlands - Erik Hitters.................... 58 What is coolness? Deconstructing and contextualizing a buzzword - Janna Michael ...................................... 59 Session #10 .................................................................................................................................................... 60 New media prosumption in the global south - Payal Arora and Nimmi Rangaswamy ..................................... 60 Watching TV fiction in the age of digitization - Nele Simons ............................................................................ 61 Mediagebruik, body dissatisfaction en appearance anxiety: een studie bij negen- tot twaalfjarige meisjes Saskia Merre and Steven Eggermont ................................................................................................................ 62 “Free your …” “most open” Android: Critical analysis of discourses on Android - Lela Mosemghvdlishvili and Jeroen Jansz ...................................................................................................................................................... 63
4
5
ETMAAL 2013
BOOK OF ABSTRACTS
Explaining globalization in pop music charts. A multilevel analysis of nine countries between 1960 and 2010 Marc Verboord and Amanda Brandellero ........................................................................................................ 65 The discourse of regional popular music (re)considered - Simone Driessen .................................................... 66 Grunting alone: online gender inequality in extreme metal music - Julian Schaap and Pauwke Berkers ........ 67 Remembering the long sixties: a European popular music heritage? - Arno van der Hoeven ......................... 68
HEALTH COMMUNICATION ........................................................................................................ 70 Session #1 ...................................................................................................................................................... 70 Health anxiety, medical consult satisfaction and the role of internet information - Martin Tanis .................. 70 Health information search of native-Dutch and Turkish-Dutch GP patients and its relation with patient participation - Sanne Schinkel, Julia van Weert, Jorrit Kester and Barbara Schouten ...................................... 71 Tailoring online health materials to improve message processing and comprehension: what strategy works best for people with low health literacy? - Corine S. Meppelink, Edith G. Smit, Bianca M. Buurman and Julia C.M. Van Weert ................................................................................................................................................ 72 Mood and time influence defensive affect regulation following self-threatening health information: evidence at the implicit level - Enny Das and Jonathan van 'T Riet ................................................................................. 74 Session #2 ...................................................................................................................................................... 75 Communicatie over gewichtige zaken met de praktijkondersteuner in de huisartsenpraktijk - Sonja van Dillen, Gert Jan Hiddink, Janneke Noordman and Sandra van Dulmen ............................................................ 75 Transmural palliative care by means of teleconsultation: a window of opportunities …and new restrictions Jelle Van Gurp, Martine Van Selm, Evert Van Leeuwen and Jeroen Hasselaar ................................................ 77 “Is echt belangrijk naar de huisarts te gaan als het aanhoudt. Vanuit anesthesiologie kunnen we hier niets mee.” De communicatieve uitdagingen van Twitterspreekuren voor gezondheidsprofessionals - Joyce Lamerichs and Charlotte Van Hooijdonk .......................................................................................................... 78 Kennis over chemotherapie: een explorerend onderzoek naar factoren die van invloed zijn op kennis van patiënten en niet-patiënten - Julia van Weert and Sifra Bolle ......................................................................... 80 Session #3 ...................................................................................................................................................... 81 Swine Flu in the E.R.: the relationship between media attention for the swine flu and registration in an emergency medicine unit - Sara Nelissen, Kathleen Beullens, Marc Sabbe and Jan Van Den Bulck ................ 81 Suffering as a discipline? Scholarly accounts on the current and future state of research on media and suffering - Stijn Joye .......................................................................................................................................... 83 E-mail versus chat counseling: a comparison of problem presentations - Wyke Stommel and Fleur Van der Houwen............................................................................................................................................................. 84 Cost-effectiveness of adherence enhancing interventions: a systematic review - Edwin Oberjé, Marijn de Bruin and Silvia Evers ........................................................................................................................................ 85 Session #4 ...................................................................................................................................................... 87 What do prenatal and postnatal women talk to their healthcare providers: findings from the Los Angeles Mommy and Baby (LAMB) survey - Yuping Mao and Lu Shi ............................................................................. 87 Put yourself in the patient’s shoes: the effect of modality and communication style on website satisfaction and recall of online health information - Nadine Bol, Ellen M.A. Smets, Hanneke C.J.M. de Haes, Eugène F. Loos and Julia C.M. van Weert.......................................................................................................................... 88 Communication about chemotherapy treatment decisions: Taking perceptions about treatment consequences into account - Sifra Bolle and Julia Van Weert .......................................................................... 90 Sewing the seeds of motherhood: the socio-political dynamics of sewing mummy blogs - Jiska Engelbert and An Kuppens ....................................................................................................................................................... 91
5
6
ETMAAL 2013
BOOK OF ABSTRACTS
Session #5 ...................................................................................................................................................... 92 Mobile fitness apps for promoting physical activity on Twitter: the #Runkeeper case - Jeroen Stragier and Peter Mechant .................................................................................................................................................. 92 A body image app: development and evaluation of a mobile app for body image improvement among adolescent girls - Jolanda Veldhuis, Elly A. Konijn and Marloes Spekman ....................................................... 93 When computer-mediated support group participation backfires: effects of pessimistic social comparisons and lack of peer support - Anika Batenburg and Enny Das .............................................................................. 94 Dating a robot is like dating a man - a new perspective on the Turing Test - Matthijs Pontier, Johan Hoorn, Elly Konijn and Rens van de Schoot .................................................................................................................. 96 Session #6 ...................................................................................................................................................... 98 Influencing the conversation: emotions, conversational valence, and binge drinking - Hanneke Hendriks, Bas Van Den Putte and Gert-Jan De Bruijn ............................................................................................................. 98 Intervening or interfering? The influence of positive and negative injunctive and descriptive norms messages protective behaviors in alcohol consumption contexts - Saar Mollen, Rajiv N. Rimal, Robert A. C. Ruiter, Su Ahn Jang and Gerjo Kok .................................................................................................................................... 99 Self-affirmation before exposure to persuasive health communications promotes anticipated regret and health behavior change - Guido M. van Koningsbruggen, Peter R. Harris and Anjes J. Smits ........................ 101 Flossing, implicit attitudes and message framing effectiveness - Gert-Jan de Bruijn, Bas Van Den Putte and Jonathan van 'T Riet ........................................................................................................................................ 102 Session #7 .................................................................................................................................................... 103 Cyberbullying on social network sites: an experimental study into bystanders’ behavioral intentions to help the victim or reinforce the bully - Sara Bastiaensens, Ilse De Bourdeaudhuij, Ann Desmet, Karolien Poels, Katrien Van Cleemput and Heidi Vandebosch ................................................................................................ 103 Modelling proximal determinants of cyberbullying perpetration - Sara Pabian and Heidi Vandebosch ....... 105 The effect of playing advergames that promote energy-dense snacks or fruit on actual food intake among children - Frans Folkvord, Doeschka Anschutz, Moniek Buijzen and Patti Valkenburg .................................. 105 Snoepen voor de televisie: Zelfbeheersing neutraliseert effect televisiereclame - Henk Westerik, Luc Kipmulder, Hans Lemmens, Ruben Konig, Esther Rozendaal and Moniek Buijzen ........................................ 106
JOURNALISM ..................................................................................................................................107 Session #1 .................................................................................................................................................... 107 ‘Facebook, daar moeten we iets mee’ Interactie tussen nieuwsmedia en het publiek op Facebook - Sanne Hille ................................................................................................................................................................. 107 The medium is the comment? De invloed van het nieuwsplatform op de lezerscommentaren - Fanny Maenhout and Michael Opgenhaffen ............................................................................................................ 108 Regionale media, democratie en burgerschap - Niek Hietbrink and Liesbeth Hermans ................................ 110 Gatekeeper of curator? De rol van sociale media in de berichtgeving over de Arabische Lente in de tvjournaals van VRT en VTM - Rebeca De Dobbelaer, Steve Paulussen and Pieter Maeseele .......................... 111 Session #2 .................................................................................................................................................... 112 Digital ambulance chasers, blogging vets and non-stop commenters. Hyperlocal news in practice - Piet Bakker ............................................................................................................................................................. 112 News ecosystem compared in two Dutch cities - Marco van Kerkhoven and Piet Bakker ............................. 114 Transparency: what newsrooms (could) do. Results of an online survey among European journalists - Huub Evers and Harmen Groenhart ......................................................................................................................... 115 Facing mediated distant suffering: a first step towards an empirical framework - Johannes von Engelhardt and Jeroen Jansz ............................................................................................................................................. 116
6
7
ETMAAL 2013
BOOK OF ABSTRACTS
Session #3 .................................................................................................................................................... 118 Politici aan tafel - de invloed van programmaformats op politieke interviews tijdens verkiezingen - Birte Schohaus and Marcel Broersma ..................................................................................................................... 118 Ideologisch pluralisme in Vlaamse nieuwsmedia: de Belgische regeringsformatie 2010-2011 - Daniëlle Raeijmaekers and Pieter Maeseele ................................................................................................................ 119 Who has got the power? A social network approach to the relationship between politicians and journalists Annemarie Walter and Willem Robert Van Hage ........................................................................................... 121 Gerepeteerde toneelstukjes of geïmproviseerde verkiezingsdebatten? Een onderzoek naar de onderhandelingen tussen politici en media tijdens de Nederlandse verkiezingscampagne van 2012 - Yael de Haan, Harmen Groenhart, Ad van Liempt, Niek Hietbrink, Bert Mateboer, Anouk Kragtwijk, Freek Diephuis, Tijmen de Groen, Jimmy Leenders, Erik Wolters and Arjanne Aleman .......................................................... 122 Session #4 .................................................................................................................................................... 123 Neoliberalism and the marketing of misery: news from Chile’s “largest ghetto” - Isabel Awad.................... 123 Klimaatverandering als definitiestrijd: een onderzoek naar de beeldvorming door maatschappelijke actoren Yves Pepermans and Pieter Maeseele ............................................................................................................ 125 Mediating travel in elite and popular newspapers: media inspired imagination of Turkey - Nilyufer HamidTurksoy and Giselinde Kuipers ........................................................................................................................ 125 Uit onderzoek blijkt. Journalistieke omgang met wetenschappelijke informatie - Chris Van Der Heijden .... 126 Session #5 .................................................................................................................................................... 128 Meer lokaal nieuwsaanbod, meer van hetzelfde nieuws - Quint Kik, Piet Bakker, Laura Buijs and Judith Katz ........................................................................................................................................................................ 128 Van boete naar beloning. Prioriteitsverschuiving in publieksverantwoording door de journalistiek - Harmen Groenhart ....................................................................................................................................................... 129 “Kwestie van gezond verstand!?” Een studie naar de zin en onzin van sociale mediarichtlijnen bij Vlaamse journalisten - Harald Scheerlinck and Michael Opgenhaffen ......................................................................... 131 Session #6 .................................................................................................................................................... 132 Freelancers in de Vlaamse journalistiek: vrije vogels of vogels voor de kat? Kwalitatief onderzoek naar randvoorwaarden freelancestatuut - Hedwig de Smaele and Rozane De Cock ............................................. 132 Professionele identiteit van journalisten werkzaam in verschillende mediatypen - Liesbeth Hermans ........ 134 Foreign correspondents in The Netherlands - Bernadette Kester .................................................................. 136 'Journalistiek in Vlaanderen door een roze bril' Een verkennend onderzoek naar genderongelijkheid in de journalistieke sector en beroepstevredenheid bij vrouwelijke journalisten - Sara De Vuyst and Karin Raeymaeckers ................................................................................................................................................. 137 Session #7 .................................................................................................................................................... 138 Prikkelend televisienieuws en de gevolgen voor de kijker: over aandacht vasthouden, informeren en betrouwbaar overkomen - Mariska Kleemans ............................................................................................... 138 Een kwestie van vertrouwen: de constructie van de geloofwaardige verteller in het journalistieke boek Gerard Smit ..................................................................................................................................................... 140 Dutch journalism practices in the digital age - Bob Nieman, Max Kemman, Martijn Kleppe and Henri Beunders ......................................................................................................................................................... 141 Hypertext en direct speech in online nieuwsartikelen. Het vergroten van de ervaren controle en inleving in het verhaal - Daniël Verheij and Luuk Lagerwerf............................................................................................ 142 Session #8 .................................................................................................................................................... 143 Shifts in Flemish cultural journalism: an analysis of historical transformations in the coverage of culture in Flemish elite and popular newspapers (1955-2010) - Sil Tambuyzer ............................................................. 143
7
8
ETMAAL 2013
BOOK OF ABSTRACTS
Strijd om het klimaat: de beeldvorming over klimaatverandering tijdens de klimaattop te Cancún in Vlaamse kwaliteitskranten - Pieter Maeseele, Yves Pepermans, Daniëlle Raeijmaekers, Laurens Van der Steen and Stijn Joye ......................................................................................................................................................... 144 Aandacht voor suïcide in Vlaamse en Waalse kranten. Imitatierisicoscores en naleving van richtlijnen voor journalisten - Jolien Vangeel and Rozane De Cock ......................................................................................... 145 Vlaamse mediaverslaggeving over cyberpesten - Anne Vermeulen and Heidi Vandebosch .......................... 146 Session #9 .................................................................................................................................................... 148 Diversiteit in het Vlaamse perslandschap: domineren Belgische volwassen mannen als elitebronnen in de hardere nieuwsthema’s? - Hanne Vandenberghe, Leen D'Haenens and Baldwin Van Gorp ......................... 148 Actor diversity in Flemish economic television news - Anoeska Schipper, Hilde Van Den Bulck and Stefaan Walgrave ......................................................................................................................................................... 150 'De documentaire is dood, lang leve de documentaire.' De Vlaamse onafhankelijke documentairesector op zoek naar een nieuw elan? - Doruntina Islamaj.............................................................................................. 151
ORGANIZATIONAL COMMUNICATION ..................................................................................153 Session #1 .................................................................................................................................................... 153 Dealing with differences: the impact of diversity perceptions on assessment of minority employees - Joep Hofhuis, Karen van der Zee and Sabine Otten ................................................................................................ 153 Engaging employees in corporate citizenship activities: the case of IBM India - Vidhi Chaudhri and Asha Kaul ........................................................................................................................................................................ 154 Great expectations: the role of mentoring schemas in protégés' and mentors' sensemaking of formal mentoring relationships - Suzanne Janssen, Marleen Mensink and Mark van Vuuren .................................. 155 Audience building by police officers on Twitter: do followers follow for business or pleasure? - Bob van de Velde and Ivar Vermeulen .............................................................................................................................. 156 Session #2 .................................................................................................................................................... 157 Medewerkers als merkambassadeurs - Joost Verhoeven and Piet Verhoeven .............................................. 157 Hoe creëer je merktrouw? Een confrontatie tussen visies uit wetenschap en praktijk - Suzanne de Bakker and Peeter W.J. Verlegh ........................................................................................................................................ 158 Het 7I netwerkmodel – 7 kritische succesfactoren voor een 3.0 organisatie - Renee Van Os, Janneke Delisse, Marco Derksen and Jan Jurriëns ..................................................................................................................... 159 'Nu moet nog blijken wat ze gaan doen. Maar het voelt positief!' Onderzoek naar de bijdrage van bewonersbijeenkomsten aan het contact tussen bewoners en de gemeente Almere - Christine Bleijenberg, Nick Dessing and Reint Jan Renes ................................................................................................................... 161 Session #3 .................................................................................................................................................... 161 The communication of emotions by organizations in times of crisis: the effect of crisis response strategy and communicated emotions on corporate reputation - Toni G.L.A. van der Meer and Joost W.M. Verhoeven 161 Negatieve berichtgeving over één bedrijf benadeelt de reputatie van de hele bedrijfstak - Jos Hornikx and Nicole van der Steen ....................................................................................................................................... 162 Hoe een organisatiecrisis wordt geframed door organisatie en pers: Een casestudie over de BP-olieramp in de Golf van Mexico - Julia Krauwer and Pytrik Schafraad .............................................................................. 164 De Communicatieprofessional als boundaryspanner; een studie naar de omgevingscommunicatie rondom de aanleg van de Noord/Zuidlijn in Amsterdam - Noelle Aarts and Mieke Muijres ............................................ 165 Session #4 .................................................................................................................................................... 167 The changing organization of humanitarian information provision: challenges of technology-mediated citizen and volunteer participation - Annemijn van Gorp .......................................................................................... 167
8
9
ETMAAL 2013
BOOK OF ABSTRACTS
New ways of working, vitality, and employee performance: the role of self determination - Claartje Ter Hoeven, Frederique Smit and Marieke Fransen ............................................................................................. 168 One size fits all? Advantages and challenges of New Ways of Working: The role of participative leadership Pernill Van der Rijt, Claartje Ter Hoeven, Joep Hofhuis and Martijn Vlug-Mahabali ..................................... 169 Session #5 .................................................................................................................................................... 170 Information sharing as strategic behaviour: the role of social motivation, time pressure and information display - Nicoleta Bălău and Sonja Utz............................................................................................................ 170 Do media matter in organizational identification? The roles of media usage and perception of communication in explaining employee identification - D.J. Schutten, J. Bartels and J.W. Ouwerkerk ......... 172 Professionalization of organizational communication by means of decision support systems - Maarten C.A. Van der Sanden, Koen H. Van Dam and Jeroen Stragier ................................................................................ 173 Reminders of death influence attitude, intention and behavior in group formation. A terror management perspective on joining a new group - Jolien Arendsen and Jessica Schutten ................................................. 175
PERSUASIVE COMMUNICATION ..............................................................................................177 Session #1 .................................................................................................................................................... 177 Does sponsorship disclosure timing influence viewers’ processing of program content? - Sophie Boerman, Eva Van Reijmersdal and Peter Neijens .......................................................................................................... 177 Sponsoring good causes: does the level of brand/cause fit matter? - Simon Zebregs, Bas van Den Putte and Marjolein Moorman ....................................................................................................................................... 178 A negation bias in word of mouth: how negations reveal and maintain expectations about brands and products - Peeter W.J. Verlegh, Camiel J. Beukeboom and Christian F. Burgers ........................................... 179 How often do you really watch? Comparing self-reported TV exposure to people-meter data - Anke Wonneberger and Mariana Irazoqui .............................................................................................................. 181 Session #2 .................................................................................................................................................... 183 Butt-naked advertisements: promoting taboo products by means of nudity - Jiska Eelen, Marieke L. Fransen and Mike Friedman ......................................................................................................................................... 183 "Elevation!" (only if you care): examining the determinants of users' elevation responses to short film clips Allison Eden and Tilo Hartmann ..................................................................................................................... 183 Increasing civic participation using 140 characters? An online experiment on the effects of different communication strategies and the use of hashtags - Mirjam Lasthuizen, Sonja Utz and Sanne Gaastra ...... 184 How influential consumers create brand-related user-generated content: The interplay between personality characteristics and motivations - Daan G. Muntinga, Marjolein Moorman, Edith G. Smit and Peeter W.J. Verlegh............................................................................................................................................................ 186 Session #3 .................................................................................................................................................... 187 Disentangling the role of brand-to-user vs. user-to-system interactivity on Facebook fanpages - Sonja Utz and Mariette Van Tuyl ........................................................................................................................................... 187 Getting too personal? Adolescents’ responses to personalized social network site advertising - Michel Walrave, Karolien Poels, Marjolijn Antheunis, Guda Van Noort and Evert Van Den Broeck ......................... 189 In-game advertising in context: the impact of social setting and brand type on the effectiveness of in-game advertising - Laura Herrewijn and Karolien Poels ........................................................................................... 190 The myth of e-mail personalization and interactivity - Remco de Vries, Tom Bakker, Damian Trilling and Piet Bakker ............................................................................................................................................................. 192 Session #4 .................................................................................................................................................... 193
9
10
ETMAAL 2013
BOOK OF ABSTRACTS
Mindless resistance to influence: when low self-control decreases yielding to persuasive communication Loes Janssen and Bob M. Fennis ..................................................................................................................... 193 Is being bad as bad as being guilty? The role of identification and just endings in evoking emotions in narrative persuasion - Hans Hoeken............................................................................................................... 195 Development of a scale to measure strategies to resist persuasive communication - Marieke Fransen, Peeter Verlegh and Claartje Ter Hoeven .................................................................................................................... 196 Implicit persuasion: How nodding the head affects people’s own attitudes - Annemarie Wennekers ......... 197 Session #5 .................................................................................................................................................... 199 The roles of brand trust and brand equity in green buying behavior: nationally certified labels versus food retail brands - Jos Bartels, Machiel Reinders and Karen Hoogendam ............................................................ 199 Explaining the impact of character-product congruence in healthy food promoting picture books - Simone de Droog, Moniek Buijzen and Patti Valkenburg ................................................................................................. 200 “You can make a difference!” Applying customization to fundraising messages - Ewa Maslowska, Edith Smit and Bas van Den Putte .................................................................................................................................... 200 Ik heb duurzame kleding gekocht: mag ik nu met de auto naar de supermarkt? - Marijn H. C. Meijers, Marret K. Noordewier, Edith G. Smit and Peeter W. J. Verlegh.................................................................................. 202 Session #6 .................................................................................................................................................... 204 The irony bias: how verbal irony reflects and maintains stereotypic expectancies - Christian Burgers and Camiel Beukeboom ......................................................................................................................................... 204 Communicating brands with auditory cues: influence brand choice with brand-related sounds - Gaby Schellekens and Peeter W.J. Verlegh .............................................................................................................. 205 "Background Noise". A study on the processing of radio advertising while media multitasking - Wendalin van de Giessen, Edith Smit and Hilde Voorveld .................................................................................................... 207 Positieve en negatieve uitingen: een krachtstrijd - Christine Liebrecht, Lettica Hustinx and Margot van Mulken ............................................................................................................................................................ 208 Session #7 .................................................................................................................................................... 209 Disclosing online behavioral advertising: effects of the cookie-icon - Guda Van Noort, Edith Smit and Hilde Voorveld ......................................................................................................................................................... 209 Protect yourself… if you can: is protection against subliminal advertising possible? - Stefan F. Bernritter, Paul Ketelaar, Thijs Verwijmeren, Johan C. Karremans, Wolfgang Stroebe and Daniël H. J. Wigboldus ............... 211 Online behavioural advertising: do we know how to cope? - Edith Smit, Hilde Voorveld and Guda Van Noort ........................................................................................................................................................................ 211 Differences between younger and older adults in responses to emotional television commercials - Margot Van Der Goot, Eva Van Reijmersdal and Mariska Kleemans .......................................................................... 213 Session #8 .................................................................................................................................................... 214 Children’s responses to advertising in social games: persuasion knowledge and susceptibility to peer influence - Esther Rozendaal, Eva Van Reijmersdal, Noortje Slot and Moniek Buijzen .................................. 214 Effects of profile targeting on children’s responses to online advertising - Eva van Reijmersdal, Esther Rozendaal, Nadia Smink, Guda van Noort and Moniek Buijzen ..................................................................... 216 Capturing children’s advertising exposure: a comparison of methods and measurements - Suzanna J. Opree, Moniek Buijzen, Eva Van Reijmersdal and Patti M. Valkenburg ..................................................................... 217 Engaging entertainment-education: how engagement in entertainment-education narratives relates to unfavorable thoughts and effects - Lonneke Van Leeuwen, Bas Van Den Putte, Reint Jan Renes and Cees Leeuwis ........................................................................................................................................................... 218 Session #9 .................................................................................................................................................... 220
10
11
ETMAAL 2013
BOOK OF ABSTRACTS
Credibility in e-WOM: how review perceptions impact their persuasiveness - Natalie Van Hemelen, Peeter Verlegh and Tim Smits .................................................................................................................................... 220 Setting the record straight: online review intentions as a function of dining experience and prior expectations - Jeana Frost, Ivar Vermeulen and Romy van der Linden .......................................................... 221 De invloed van sociale en interpersoonlijke identificatie op de beoordeling van online reviews - Alexander Schouten and Marjolijn Antheunis ................................................................................................................. 222 This T-shirt is A-W-E-S-O-M-E!: the creation of camp products through ironic online reviews - Camiel Beukeboom, Christian Burgers and Amber Scholtz ........................................................................................ 223 Session #10 .................................................................................................................................................. 225 Evolutionaire rollen en commerciële communicatie naar ouderen - Dirk Sikkel ........................................... 225 Naar een contextueel begrip van digitale uitsluiting: de rol van sociaal kapitaal in ICT-toegang en –gebruik van vrouwen van middelbare leeftijd - Dana Schurmans and Ilse Marien ..................................................... 226 Webcare als online reputatiemanagement: het gebruik van webcarestrategieën en conversational human voice in Nederland, en de effecten hiervan op reputatie - Judith Huibers and Joost Verhoeven .................. 227 De kip of het ei?: Facebook & merkervaring - Harm Ter Halle and Joost Verhoeven .................................... 228
POLITICAL COMMUNICATION .................................................................................................231 Session #1 .................................................................................................................................................... 231 Predicting participation: personality, political attitudes and the use of participatory media - Tom P. Bakker, Klaus Schoenbach and Claes De Vreese ......................................................................................................... 231 Spiral of political learning. The reciprocal relationship of news media use and political knowledge - Judith Moeller ........................................................................................................................................................... 232 Internetgebruik en betrokkenheid bij de politiek - Marjon Schols, Jos de Haan and Jeroen Jansz ................ 232 How debates, the news and voting advice applications add to issue ownership in the explanation of shifts in party preferences: the 2012 Dutch election campaign - Jan Kleinnijenhuis, Nel Ruigrok, Anita Van Hoof and Wouter van Atteveldt ..................................................................................................................................... 233 Session #2 .................................................................................................................................................... 235 Media and/ or Reality: Mediated Information, a replacement for, or an addition to Real World Developments as a Source of Anti-immigrant Sentiment in the Netherlands, Denmark, and Sweden, 2003-2010? - Marijn Van Klingeren, Hajo G. Boomgaarden, Rens Viegenthart and Claes de Vreese ............................................. 235 Media coverage of Muslims: the foreign threat versus the domestic other - Anouk Van Drunen ................ 236 Gebruik van sociale media en deelname aan de Egyptische opstand: de rol van toekomstverwachtingen, democratische attitudes en vertrouwen in de overheid - Peter Kerkhof, Andre Krouwel, Jacquelien van Stekelenburg and Bert Klandermans .............................................................................................................. 237 Are Twitter and YouTube the new networked news wires? A quantitative content analysis of journalists’ sourcing practices during the Arab Spring - Sarah Van Leuven, Annelore Deprez and Karin Raeymaeckers . 238 Session #3 .................................................................................................................................................... 239 Quality of public debates - Regula Hänggli and Richard van der Wurff .......................................................... 239 It’s funny – but is it appropriate? Political humor in the media and its conditional impact on citizens’ social trust and efficacy - Sophie Lecheler and Andreas Schuck .............................................................................. 239 Style Talk. An experimental investigation into direct effects of non-traditional political communication styles on the perceived credibility of politicians - Niels Bouwman and Linda Bos ................................................... 241 The Plain, familiar and comfortable language of the common people. An examination of rhetorical figures in populist political language - Lieuwe Kalkhoven and Christ'L De Landtsheer .................................................. 242 Session #4 .................................................................................................................................................... 243
11
12
ETMAAL 2013
BOOK OF ABSTRACTS
How political candidates use Twitter and the impact on Votes - Sanne Kruikemeier, Guda Van Noort, Rens Vliegenthart and Claes De Vreese .................................................................................................................. 243 Evelien D'Heer, Pieter Verdegem and Peter Mechant: Tweeting the elections? An analysis of Twitter use during the 2012 local elections in Flanders .................................................................................................... 245 Heidi Schulze and Damian Trilling: Of binders and bayonets: measuring the adoption of frames from the US presidential debates on Twitter ..................................................................................................................... 246 Session #5 .................................................................................................................................................... 247 Gaining attention but communicating little? On the conditionality of episodic and thematic coverage of protest events - Ruud Jesse Wouters ............................................................................................................. 247 Media populism over time - Linda Bos and Kees Brants ................................................................................ 248 Directe aandacht voor retoriek en negativiteit - Luuk Lagerwerf, Charlotte Brussee and Amber Boeynaems ........................................................................................................................................................................ 250 Media cues and support for right-wing populist parties - Penelope Sheets, Linda Bos and Hajo Boomgaarden ........................................................................................................................................................................ 251 Session #6 .................................................................................................................................................... 252 Framing climate change - emotional responses as mediators for the effect of threat framing on individual risk perceptions and behavioral intentions - Andreas Schuck and Lukas Otto ..................................................... 252 Human interest framed television news and its effect on support for government intervention - Mark Boukes, Hajo G. Boomgaarden, Marjolein Moorman and Claes H. de Vreese ............................................... 253 Teaching the computer to code frames in news: a supervised machine learning approach to frame analysis Bjorn Burscher, Daan Odijk, Rens Vliegenthart, Maarten de Rijke and Claes de Vreese ............................... 255 Frame dynamics and complexity in the financial crisis - Dirk Oegema, Friederike Schultz, Jan Kleinnijenhuis, Wouter Atteveldt and Sabine Geers ............................................................................................................... 256 Session #7 .................................................................................................................................................... 257 Which politicians pass the media gates and why? A preliminary model and review of previous studies Debby Vos ....................................................................................................................................................... 257 When politics becomes news: an analysis of parliamentary questions and press coverage in three WestEuropean countries - Rosa Van Santen, Peter Van Aelst and Luzia Helfer ..................................................... 259 When do proposed pension reforms lead to a media wave? - Roel Dijkman, Wouter Van Atteveldt, Barbara Vis, Laura Hollink and Pim van de Pavoordt ................................................................................................... 260 Absent yet popular? Explaining news visibility of Members of the European Parliament - Katjana Gattermann and Sofia Vasilopoulou ................................................................................................................................... 261
12
13
ETMAAL 2013
BOOK OF ABSTRACTS
Abstracts Etmaal 2013
13
14
ETMAAL 2013
BOOK OF ABSTRACTS
CULTURE & ENTERTAINMENT Session #1 Thursday 7 February 14:05 – 15:05 hrs. Chair: Anouk van Drunen Location: Santander (M2-11) English session
ADOLESCENTS’ INSTRUMENTAL AND EXPRESSIVE TEXT MESSAGING BEHAVIOR AND THEIR FRIENDSHIP QUALITY: A PERCEIVED AFFORDANCES-APPROACH - MARIEK VANDEN ABEELE, ALEXANDER SCHOUTEN AND MARJOLIJN ANTHEUNIS The mobile phone is fully integrated into adolescents’ daily lives. A reason for the popularity of mobile phones is that they aid identity development by affording teens greater autonomy in the support of their relationships by providing them with a private communication channel with which they can communicate with their friends (and family) anytime and anyplace. Research has shown that teen-to-teen texting comprises the bulk of mobile communications that teens engage in. The content of these text messages ranges from highly instrumental messages for micro-coordinating teens’ everyday logistics, over expressive messages with little emotional content (e.g., making jokes, chatting), to expressive messages with a more socio-emotional content (e.g., intimate self-disclosure). With respect to the latter contents, text messaging may be particularly attractive to teens as its text-based, a-synchronous nature allows them to exert greater control over their writings. Hence, the current study wishes to contribute to the field of adolescent mobile phone research by exploring (1) how teens’ perceptions of three affordances of mobile communication technology (having a private communication channel, anytime anyplace availability, and controllability over one’s messages) predict the use of text messaging for micro-coordination (instrumental), for passing the time (expressive, low socio-emotional content) and for self-disclosing intimate information (expressive, high socio-emotional content), and (2) how these behaviors, in turn, affect two characteristics of relationship quality, namely friend companionship and friend support. With respect to the relationship between perceived affordances and text messaging behavior, we hypothesize that adolescents who value the mobile phone more strongly for its ‘private communication’-character and for its ability for anytime/anyplace interaction will be more likely to engage in expressive text messaging behaviors (H1a, H1b); that adolescents who value anytimeanyplace accessibility more strongly will be more likely to engage in micro-coordination (H2a) and texting to pass the time (H2b); and that adolescents who value the controllability over one’s writings more strongly, will be more likely to engage in texting for intimate self-disclosure (H3a). With respect to the relationship between text messaging behavior and relationship quality, we expect to find positive relationships between all variables, but particularly strong relationship between micro14
15
ETMAAL 2013
BOOK OF ABSTRACTS
coordination and (face-to-face) companionship (H4) and between intimate self-disclosure and support (H5). We investigated our hypothesis by conducting a quantitative survey among 1943 (50.6% males,Mage = 15.28, SD = 1.89). Results generally support our hypotheses. First, adolescents who perceive text messaging as being able to have a private communication channel more often use text messaging for both passing time and intimate self-disclosure, confirming H1a and H1b. Second, the ability for anytime/anyplace communication was positively related to both using texting for passing time and micro-coordination (H2a and H2b). Third, controllability was positively related to intimate selfdisclosure (H3) but also to using text messaging for micro-coordination. Finally, we indeed confirmed that using texting for micro-coordination was mostly related to face-to-face companionship (H4), and using texting for intimate self-disclosure was most strongly related to receiving social support. Finally, unexpectedly, we found that using text messaging to pass time was also positively related to face-to-face companionship. Our results indicate that the mobile phone affordances perceived by teens significantly predict to what ends they use the mobile phone, and that these uses, in turn, predict the companionship teens experience from their friends and how strongly they can turn to their friends for emotional support. A parallel can be drawn here between the findings of our study and the findings from previous research on teens’ internet use and their social relationships (e.g., the role of perceived CMC attributes in the internet-enhanced self-disclosure hypothesis). The findings of our study also point towards the relevance of the perceived affordances-concept as a (cognitive) psychological antecedent of mobile phone use. The concept of perceived affordances may be examined further in future research by exploring how particular inter- and intra-personal characteristics may affect differences in perceived affordances.
DEVELOPMENT AND VALIDATION OF THE ADOLESCENT MEASURE OF EMPATHY AND SYMPATHY (AMES) - HELEN VOSSEN, JESSICA PIOTROWSKI AND PATTI VALKENBURG Empathy is an important concept within communication research. In persuasive communication research, empathy is seen a key mediator of persuasive effects. Empathy can have a positive effect on persuasion by increasing identification (Cohen, 2006) or by inhibiting psychological reactance (Shen, 2011). Furthermore, media-effects research has shown that playing violent video games decreases empathy (Anderson, Shibuya et al. 2010), which in turn is related to increases in aggression (Jolliffe & Farrington, 2004). Despite the importance of empathy, valid and reliable measures are scarce, especially for adolescents. To address this gap in methodological literature, this study reports the development and validation of a self-report measure titled “The Adolescent Measure of Empathy and Sympathy (AMES)”. Throughout the years, different definitions of empathy have been used. Initially, researchers emphasized the affective nature of empathy and defined it as a vicarious emotional response to the perceived emotional experiences of others or the experience of another’s emotional state (e.g., Mehrabian & Epstein, 1972; Stotland, 1969). More recently, researchers have come to recognize that empathy is not only defined as the experience of another person’s emotion but also the 15
16
ETMAAL 2013
BOOK OF ABSTRACTS
comprehension of another person’s emotion, which can be considered the cognitive component of empathy (e.g., Feshbach, 1997; Hoffman, 2001). Although definitions differ in their emphasis on the cognitive or affective components of empathy, the general consensus is that empathy encompasses both components. Given that empathy plays a key role in several developmental milestones for children and adolescents including moral development (Eisenberg, 2000), and prosocial behaviour (Roberts & Strayer, 1996), it is not surprising that developmental psychologists have developed many instruments to assess empathy in adolescents and children. Some examples include the ‘Index of Empathy for Children and Adolescents’ by Bryant (1982), the empathy subscale from the ‘Children’s Behaviour Questionnaire’ (CBQ; (Rothbart, Ahadi, & Hershey, 1994), the ‘Interpersonal Reactivity Index’ (Davis, 1980) and the ‘Basic Empathy Scale’ (Jolliffe & Farrington, 2006). Although these scales are widely used, there are at least three critical limitations to these measures. First, although some scales distinguish between an affective component and a cognitive component of empathy, there are others that measure empathy as a single construct (e.g. CBQ) or that calculate a total score based on subscales that do not correlate positively (IRI). Second, although empathy and sympathy are distinct and separable concepts, several empathy scales equate empathy with sympathy (e.g. CBQ). Relatedly, in several scales (e.g. Index of Empathy for Children and Adolescents’), empathy is confused with constructs such as ‘empathic sadness’ (de Wied et al., 2007) or perspective taking (Davis, 1980). Finally, in some scales (e.g. IRI), the wording of the items is ambiguous which leaves too much room for differences in interpretation. This ambiguity is particularly problematic adolescents (de Leeuw, Borgers et al. 2004). The AMES was created to address these limitations. Specifically, the AMES 1) balances the emphasis on affective empathy and cognitive empathy, 2) distinguishes between empathy and sympathy by including a specific sympathy subscale, and 3) uses unambiguous wording appropriate for adolescent self-report. To develop the scale, a total of nineteen items measuring affective empathy (e.g. “I feel scared when a friend is afraid” ), cognitive empathy (e.g. “I can tell when a friend is angry even if he/she tries to hide it”) and sympathy (e.g.“ I feel sorry for a friend who is sad”) were created. Adolescents aged 1015 years old (n = 499) participating in an online survey answered these 19 questions using a 5-point frequency scale (never, almost never, sometimes, often, always). The AMES had to meet the standards of reliability (i.e., internal consistency), validity (i.e., crosspopulation validity and construct validity), and utility (i.e., a reliable and valid scale with a minimum number of items). To increase the utility of the scale, the initial 19 items were reduced to 12 using exploratory factor analyses. Confirmatory factor analyses verified the intended three factor structure . The affective empathy scale, cognitive empathy scale and sympathy scale all showed acceptable internal consistency (alpha = .72 to .88). All three scales correlated positively with prosocial behaviour and negatively with aggressive behaviour. Further, in line with the existing literature (Mestre, Samper, Frias & Tur, 2009), girls were found to have significantly higher scores than boys on all three scales providing. These relationships across prosocial behaviour, aggression, and gender provide support for the construct validity of the measure. Taken as a whole, the AMES proves to be a reliable and valid measure for empathy and sympathy in adolescents. 16
17
ETMAAL 2013
BOOK OF ABSTRACTS
UNDERSTANDING THE ROLE OF PERCEIVED PEER NORMS IN THE RELATIONSHIP BETWEEN SEXUAL MEDIA USE AND YOUNG ADOLESCENTS’ SEXUAL BEHAVIORS - ELINE FRISON, LAURA VANDENBOSCH AND STEVEN EGGERMONT Popular mass media have been criticized for promoting biased sexual messages among adolescents (Brown & Steele, 1995; Taylor, 2005). In line with this criticism, sexual media have been found to contribute to adolescents’ development of adversarial peer norms regarding sexuality (Brown & Steele, 1995) and to predict an early initiation of sexual intercourse (e.g., Kinsman et al., 1998). Moreover, perceived peer norms have also been shown to predict early sexual initiation (e.g., Collazo, 2004). Several studies have warned that an early sexual debut is related to various harmful consequences, such as an increased risk for sexually transmitted diseases and teenage pregnancies (e.g., Abma & Sonenstein, 2001). The current three-wave panel study aimed to test an integrative model that organizes prior research on the relationships among sexual media exposure, perceptions of peers’ sexual activities, and sexual behavior. This study aims to improve scholars’ knowledge about this topic in three ways. First, the study responds to a call for a better understanding of the mechanisms that are at work in the relationship between sexual media exposure and sexual behavior (Collins et al., 2004). One of these mechanisms may be perceived peer norms because prior research has shown that exposure to sexual media is positively related to both perceptions of peers’ sexual activities (e.g., Taylor, 2005) and adolescents’ sexual behavior (e.g., Kinsman et al., 1998). By applying a longitudinal research design, the study aims to draw causal links among the relationships under scrutiny and to conduct a more severe test of mediation (Martino et al., 2005). Second, there is a striking need for more research among younger adolescents (e.g., Taylor, 2005), since previous studies have mainly examined the relationships between media use and sexual attitudes and behaviors in middle –and late adolescent samples. However, because of their early developing sexual identity, young adolescents are more likely to be affected by media use (Ward, 2003) and perceived peer norms (Vandenbosch & Eggermont, 2011).The present study will therefore focus on adolescents. Third, previous research has suggested that the influence of media exposure on the perceptions of peers’ sexual activities (e.g., Vandenbosch & Eggermont, 2011) may differ according to gender and has called for more research on the role of gender. An additional aim of the current study is to address this gap. To answer our research goals, a three-wave panel study with an interval of six months was conducted among 12- to 15-year old adolescents (N = 762; Mage = 14.07, SDage = .84). During three consecutive time periods, participants were asked to answer questions about their age, educational level, media exposure (i.e., music videos and men/women magazines), perceptions of male/female peers’ sexual activities and sexual behavior. Based on the collected longitudinal data, a model for boys and for girls was estimated and tested with structural equation modeling (AMOS). In line with previous research (Baumgarter, Valkenburg, & Peter, 2011), our models controlled for age and educational level at wave 1. The structural equation model for boys showed significant, but surprisingly negative relationships between exposure to music videos (time 1) and perceptions of both male and female peers’ activities (time 2). The model also showed a negative significant relationship between perceptions of male 17
18
ETMAAL 2013
BOOK OF ABSTRACTS
peers’ sexual activities (time 2) and sexual behavior (time 3). No significant relationship was found with reading magazines. Additionally, a positive direct relationship was found between exposure to music videos (time 1) and sexual behavior (time 3). The structural equation model for girls indicated that no direct relationships between media exposure (time1) and perceived peer norms (time 2) existed. Perceptions of male and female peers’ sexual activities (time 2) positively predicted sexual behavior (time 3). Again, no significant relationship was found with reading magazines. Taken together, the findings of the current longitudinal study contribute to a more complete understanding of the complex relationships between media use, perceived peer norms and sexual behavior. The study’s unexpected findings raise questions about the conclusions of previous crosssectional studies on sexual media effects. Discussion focuses on such as explanations as Heterosexual Script Theory and the False Consensus Effect and highlights the need for more longitudinal research with special attention toward the emergence of gender differences in sexual media effects.
TEEN SEXTING ACROSS EUROPE: A CROSS-NATIONAL COMPARISON OF THE PREDICTORS OF SEXTING - SUSANNE E. BAUMGARTNER, SINDY SUMTER, JOCHEN PETER AND PATTI M. VALKENBURG Sexting – the sending or posting of sexual photos or messages via electronic devices – has received considerable attention from media and researchers (Lenhart, 2009; Livingstone, Haddon, Görzig, & Olafsson; Mitchell, Finkelhor, Jones, & Wolak, 2012; The National Campaign to Prevent Teen and Unplanned Pregnancy, 2009). However, the vast majority of studies on sexting has been investigated from a national perspective, predominantly in the US. Therefore our knowledge about this behavior among European youth is highly limited. Moreover, cross-national comparisons of sexting are widely missing. As a result, contextual factors on the country level have often been ignored. However, taking the context into consideration is important for at least two reasons. First, the differences in prevalence and frequency of sexting across countries may be due to specific characteristics of a country. For example, in countries with higher internet penetration, adolescents may have more opportunities to engage in this behavior. Second, predictors of sexting may vary across countries. Whereas in some countries, specific individual predictors have a strong influence on sexting due to characteristics that these countries share, the same factors may have no or only a weak influence in other countries where the particular characteristics are absent. Investigating contextual factors next to individual factors across different countries may thus provide a more comprehensive picture of youth sexting. The present study employed data from the EU Kids Online II project, including 14,946 11 to 16 year old adolescents (49.7% boys) from 20 European countries. This dataset provided the unique opportunity to study sexting from a cross-national perspective. More specifically, the aim of this study was to investigate which factors on the individual and the contextual level explain why adolescents engage in sexting. On the individual level, we investigated age, gender, sensation seeking, and frequency of internet use. On the contextual level, we investigated traditionalism, gross domestic product (GDP), and broadband penetration. Moreover, the study investigated whether individual predictors vary across countries and whether this can be explained by characteristics of the country.
18
19
ETMAAL 2013
BOOK OF ABSTRACTS
Multilevel analysis revealed that the prevalence of sexting varied across countries (10% of the variance in sexting was due to differences between countries). In most countries, only 1% to 5% of boys and 1% to 4% of girls engaged in sexting. Only in the Czech Republic and Sweden, prevalences for sexting were much higher both for boys (10.3% and 12.9%) and girls (10.1% and 10.2%). The individual-level characteristics age, sensation seeking, and frequency of internet use predicted sexting across all countries, (Odds ratios = 1.30, 1.63, 1.01, respectively, all p < .001). The influence of gender varied across countries (variance component for gender = 0.30, df = 19, χ² = 45.40, p = .001). Although country characteristics had no direct effect on adolescent sexting, traditionalism significantly predicted gender differences in sexting (OR = 0.17, p < .05). In more traditional countries, gender differences were stronger with more boys than girls engaging in this behavior. In traditional countries, girls may be more restricted in their behavior, in particular in their sexual behavior. In these countries, girls may be less able to express their sexuality. In less traditional countries, these gender differences were less apparent or even reversed. This finding is in line with previous findings showing that gender differences in sexuality tend to be smaller in permissive societies (Petersen & Hyde, 2010). In sum, the present study shows that individual as well as cultural characteristics influence sexting among youth. When investigating sexting, the broader cultural context should, therefore, be taken into consideration.
Session #2 Thursday 7 February 15:15 – 16:15 hrs. Chair: Jeroen Jansz Location: Santander (M2-11) English session
PRACTICE MAKES PERFECT: THE LONGITUDINAL EFFECT OF ADOLESCENTS’ INSTANT MESSAGING ON THEIR ABILITY TO INITIATE OFFLINE FRIENDSHIPS - MARIA KOUTAMANIS, HELEN G.M. VOSSEN, JOCHEN PETER AND PATTI M. VALKENBURG An important developmental task in adolescence is to learn how to form and maintain close, meaningful relationships. In order to accomplish this task, adolescents need to develop social competence, which consists of at least the ability to (a) initiate friendships and other relationships, (b) appropriately present and disclose oneself to others (i.e., self-disclosure), and (c) be assertive when necessary (Valkenburg & Peter, 2008). Traditionally, social skills are developed in face-to-face communication (Valkenburg & Peter, 2011). However, with the rise of online communication the opportunities to learn and rehearse social skills have increased dramatically. Online communication stimulates adolescents to talk about intimate subjects (Joinson, 2001; Tidwell & Walther, 2002), edit and thus optimize their self-disclosure and self-presentation (Valkenburg & Peter, 2011; Walther, 2007), and communicate with a variety of people (Valkenburg & Peter, 2011). Our Internet-induced social skills hypothesis therefore argues that online communication can stimulate adolescents’ offline social skills because it offers them opportunities to rehearse these skills online (Valkenburg, Sumter, 19
20
ETMAAL 2013
BOOK OF ABSTRACTS
& Peter, 2011). Although an Internet-induced social skills hypothesis seems plausible, it has received little scholarly attention. The few studies on the relationship between online communication and adolescents’ offline social competence showed mixed results. One study found a positive relation between online communication and offline social competence (Valkenburg & Peter, 2008), whereas another did not (Harman, Hansen, Cochran, & Lindsey, 2005). In addition, the literature on both social competence and friendships has several weaknesses that prevent us from drawing decisive conclusions concerning the influence of adolescents’ online communication on their offline social competence. First, neither of the studies has assessed the effects of general online communication. Second, most are cross-sectional studies and therefore unable to single out the causality of the relationships between online communication, social competence, and friendships. Third, there is too little systematic research on possible underlying mechanisms of the relationship between online communication and offline social competence. The current study addressed the aforementioned weaknesses of previous research using a longitudinal design and focusing specifically on one crucial aspect of adolescents’ social competence: their ability to initiate offline friendships. In addition, we investigated underlying mechanisms of the relationship between online communication and offline social competence. We focused on two possible mechanisms: the opportunities offered by instant messaging (a) to communicate with a variety of people and (b) to disclose intimate information. We conducted a three-wave longitudinal study among 690 Dutch adolescents between 10 and 17 years old (50% girls). Data were collected on three occasions with six-month intervals. We measured adolescents’ use of instant messaging software (IM use), their ability to initiate offline friendships (initiation), their tendency to communicate online with people of various ages and cultural backgrounds (diversity of online communication partners), and their online self-disclosure. We tested three autoregressive cross-lagged models: (a) the direct effect of IM use on initiation, (b) the effect of IM use on initiation with the diversity of online communication partners as a mediator, and (c) the effect of IM use on initiation with online self-disclosure as a mediator. Results showed that adolescents’ IM use increased their ability to initiate offline friendships over time. Furthermore, IM use indirectly increased adolescents’ ability to initiate offline friendships through the diversity of their online communication partners. Although IM use increased adolescents’ self-disclosure in the following wave, online self-disclosure did not increase initiation. In contrast, initiation increased adolescents’ online self-disclosure in the subsequent wave. In spite of public concern over negative Internet effects, this study showed that online communication can have a positive effect on the development of adolescents’ offline social competence. More specifically, through instant messaging adolescents can practice social skills online and learn to relate to a variety of people. Over time, they may apply these skills in face-to-face situations, increasing their ability to initiate offline friendships.
HOW SOCIAL ARE SOCIAL NETWORK SITES? A LONGITUDINAL STUDY AMONG EARLY ADOLESCENTS - MARJOLIJN L. ANTHEUNIS, ALEXANDER P. SCHOUTEN AND EMIEL KRAHMER 20
21
ETMAAL 2013
BOOK OF ABSTRACTS
Social network sites (SNS) such as Facebook are deeply ingrained into our daily lives. Especially adolescents spent much time on SNS. This massive popularity of SNS among adolescents elicits worries in society. Concerns have been voiced that adolescents’ social connectedness would be reduced by the increased use of SNS (e.g., Livingstone, 2008; Turkle, 2011). However, despite the significant presence of SNS in the lives of early adolescents and the amount of discussion in the popular press about how SNS is changing social life, no studies have examined the effects of SNS in the social lives of early adolescents. Particularly for early adolescents these sites can play a significant role in their social lives. Adolescents are especially likely to turn to SNS to meet new friends and get to know each other (Hinduja & Patchin, 2008; Valkenburg, Schouten & Peter, 2005). Moreover, in early adolescence, family relationships become less important and there is a sharp increase in the number and importance of peer relationships (Harter, 1999; Hartup, 1996). This development is reflected in adolescents’ internet use. The amount of online friends sharply increases during early adolescence (Boyd, 2008; Duimel, 2009). Therefore, the aim of this study is to examine the effects of SNS in early adolescents’ social lives. In this study, we examine if SNS use (i.e., time spent on SNS, amount of friends on SNS, and connectedness with the SNS) affects early adolescents’ (a) quality of friendship, (b) bridging social capital, (c) bonding social capital, and (d) social competence. Based on theory and earlier research we expect positive effects of SNS use on friendship quality (H1), bridging social capital (H2), bonding social capital (H3), and four dimensions of social competence: ability to self-disclose (H4a), initiation of interactions (H4b), assertiveness (H4c), and supportiveness (H4d). Method We conducted a two wave panel survey1 among 1.681 adolescents between 11 and 14 years of age (M = 12.61; SD = 0.70; 57.8% girls), of which 1.322 adolescents (78.8%) were a member of an SNS (N = 1.322). The adolescents were students from 14 different secondary schools in different parts (rural and urban) of the Netherlands. After we obtained institutional approval and parental consent, the questionnaires were administered on paper in the respondents’ classrooms. Completing the questionnaire took about 20 minutes each wave. In the questionnaire the adolescents SNS use, friendship quality, bridging social capital, bonding social capital, social competence (including all four dimensions) and demographics were measured. All scales were based on prior research. All the variables loaded on a one-dimensional scale with high Cronbach’s alphas in each wave. Results and conclusion Hypotheses were tested using regression analyses1. Results in Table 1 show that almost all the hypotheses can be supported, with exception of the effect of SNS use on assertiveness (H4c) one of the four dimensions of social competence. Table 1. Effects of SNS use in wave 1 on the dependent variables in wave 2 (controlled for the dependent variable in wave 1). B SE Β Quality of friendship .06 .02 .07** 21
22
ETMAAL 2013
BOOK OF ABSTRACTS
Bridging social capital .07 .02 .08** Bonding social capital .05 .02 .06* Social competence .05 .02 .07** - Ability .10 .03 .09*** - Initiation .05 .02 .05* - Assertiveness .04 .02 .04 - Supportiveness .05 .02 .07** Note. * p < .05, ** p < .01, *** p < .001. In conclusion, the results are in line with our predictions that SNS use positively affected friendship quality, bridging social capital, bonding social capital, and social competence (the ability to selfdisclose, initiation of interactions, and supportiveness). Contrary to the concerns that adolescents’ social connectedness would be reduced by the increased use of the Internet in general and SNSs in specific (e.g., Livingstone, 2008; Nie, 2001; Turkle, 2011), the outcomes of this study suggest that there are positive social outcomes of using an SNS. After all, intensely using an SNS have positive effects on early adolescents’ friendship quality, their experience of social capital, and their social competence. All these aspects are vital in early adolescents’ daily lives and their social connection with friends. Endnote 1The results in this study are based on two waves. However, we recently conducted a third wave. The data of this third wave will be added to our dataset next month. If this study will be accepted for the conference, we present the data of all three waves, analyzed with structural equation modeling instead of regression analysis.
LINKED IN OR LEFT OUT - IMPRESSION FORMATION BASED ON ONE’S PROFESSIONAL AND PRIVATE SOCIAL NETWORK SITE PROFILE - CHRISTINA JASCHINSKI, SOMAYA BEN ALLOUCH AND ARD HEUVELMAN Social network sites (SNSs) have become an essential medium in the modern world to manage one’s private and professional relationships. However, research in this field is still explorative and a lot of questions are yet unanswered. The present research investigated if recruiters can form reliable and valid impressions of potential employees based on their social network profile. While current studies investigate impression formation based on private SNSs, similar research in the field of professional SNSs is scarce. Moreover, scientists just recently have started to look into SNSbased impression formation in a recruiting context. However, studies are still explorative and more research is needed in this field. The present study wants to address these issues by investigating both, the use of private and professional SNSs in employment selection. More specifically, a comparison between the more private-oriented SNS Facebook and the professional-oriented SNS LinkedIn is made. Following this objective, a sample of six experienced Dutch recruiters who work in local employment agencies volunteered as raters in this research. The subjects rated the personality and intelligence of 30 targets based on their Facebook or LinkedIn profile. The subjects were randomly assigned to one 22
23
ETMAAL 2013
BOOK OF ABSTRACTS
of the two conditions, respectively looking at Facebook or LinkedIn profiles. Each condition had one male and two female raters. The three recruiters in the Facebook condition were aged between 29 and 35 (M=32, SD=3.00). The three recruiters in the LinkedIn condition were a little younger with M=23.67 (SD=1.53). Thirty stimulus profile pairs were sampled among Dutch (40%) and German (60%) graduate students or recent graduates. Graduate students are considered as a suitable target group as they are about to enter the job market, where they will be confronted with companies’ assessment and screening procedures. Moreover, LinkedIn (2012) announced students and recent graduates to be their fastest growing demographic group. To sample in a time-efficient manner the snowball principle was utilized to gather the relevant profiles, asking the owner for their consent before using their profiles. As a precondition targets were obliged to own both, a LinkedIn and a Facebook profile. The final sample consisted of one third male and two third female profile owners. The participants were aged between 23 and 29 (M= 25.07, SD= 1.66). The stimulus material consisted of several screenshots. Targets’ profiles were captured in several screenshots. With regard to the Facebook condition, the relevant cues were depicted in three frames: top part of the front page, ‘About’ section and ‘Favourites’ section. Previous research has shown that these parts contain elements relevant for impression formation such as photographic material of the target and his acquaintances, number of connections, wall-posts, biographical data, identity claims and behavioural residue. The side section which encloses redundant advertising material was removed. There is little scientific knowledge about which parts of the LinkedIn profile are relevant to impression formation. Therefore, the complete public profile was captured. The objective of this research was to investigate if recruiters are able to form reliable and valid impressions of a target based on the target’s SNS-profile. To our knowledge, this research is the first one which incorporates impressions deriving from LinkedIn profiles and contrasting them with Facebook-based impressions. Consistent with our hypothesis recruiters agreed in personality impression with regard to all traits, but Agreeableness and Emotional Stability when looking at Facebook profiles. Consensus was strongest for Openness to Experience followed by Extraversion and Conscientiousness. Although not always matching the exact order, this pattern is similar to related research on SNS as well as research on personal environments. Moreover, these findings support the notion that some traits are more visible then others and that visibility of a trait is related to the agreement among observers. Extraversion seems to be highly visible across contexts (online and offline, acquaintance and zero-acquaintance, private and professional relationships) and is therefore consistently detected with a high degree of consensus. Openness to Experience, which showed the highest consensus in the present research, and Conscientiousness, which ranked third, performed similarly well in other SNS-related studies. In contrast, Emotional Stability and Agreeableness are less visible and are consistently found to have weak consensus. Based on the results it can be concluded that more scientific research is needed in the field of SNS based impression formation as results concerning the validity of these impressions are inconsistent. Furthermore, organizations should rethink the use of SNSs in employment selection, as it could lead to inaccurate assessments of applicants’ capacities.
23
24
ETMAAL 2013
BOOK OF ABSTRACTS
Session #3 Thursday 7 February 16:45 – 17:45 hrs. Chair: Patti Valkenburg Location: Santander (M2-11) English session
SEXUALIZING SITCOMS, SELF-OBJECTIFICATION AND THE SEXUAL DOUBLE STANDARD: A LONGITUDINAL STUDY AMONG ADOLESCENTS - LAURA VANDENBOSCH AND STEVEN EGGERMONT During the last decade, scholars have been increasingly concerned with the level of sexualization in contemporary Western society (Moradi & Huang, 2008). This sexualization is expected to occur both in personal interactions and in mass media (Fredrickson & Roberts, 1997). For instance, sitcoms have repeatedly been shown to emphasize the sexual appeal of characters and to value appearance ideals strongly (Kim et al., 2007; Ward, 1995). Several studies have reported that exposure to such messages may cause an experience of sexualization among media users (e.g., Aubrey, 2006; Vandenbosch & Eggermont, 2012). These sexualizing experiences have been found to trigger a three-step process that encompasses the internalization of appearance ideals, self-objectification and body surveillance (Moradi & Huang, 2008). In turn, these components have been shown to adversely affect one’s sexuality (Fredrickson & Roberts, 1997; Steer & Tiggemann, 2008). Consistent with the research tradition on patriarchal gender stereotypes (Fromme & Emihovich, 1998; Reiss, 1956), objectification scholars hypothesize increased adherence to traditional sexual gender roles as a result of self-objectification (e.g., Gillen, Lefkowitz, & Shearer, 2006). In turn, adherence to traditional sexual scripts or the so-called sexual double standard has been shown to be harmful for the developing ability of adolescents to engage in authentic, rewarding sexual interactions (Sanchez, Fetterolf, & Rudman, 2012, p.17) The current study aimed to propose and test an integrative model that organizes prior research on the relationship among exposure to sexualizing media, the three-step process of self-objectification, and gender-stereotypical sexual attitudes. This study expands prior research on this topic in three ways. First, the dominant design that has been used to examine relationships between media exposure and increased endorsement of gender-stereotypical sexual attitudes is the cross-sectional survey (Ward, 2003). Likewise, cross-sectional surveys have primarily been used to explore associations between media exposure and internalization, self-objectification and body surveillance (Grabe & Hyde, 2009; Vandenbosch & Eggermont, 2012). However, although both lines of research focus on the effects of media, studies have rarely applied a longitudinal design to test assumptions regarding temporal order (Moradi & Huang, 2008; Ward, 2003). This study aimed to address this gap by using data from a three-wave panel study (N = 1,041).
24
25
ETMAAL 2013
BOOK OF ABSTRACTS
Second, because of the profound changes that teenagers experience, scholars in this field have called for greater attention to the effects of sexualization on adolescents (e.g., APA, 2007). The extensive developments in adolescents’ body image (Markey, 2010) and sexuality (Ward, 2003) have been hypothesized to increase adolescent awareness of their appearance and sexuality and thus highlight the need for more research on how the sexuality and body image of young people may be influenced by exposure to sexualizing media. The current study therefore focused on 12- to 18-year-old adolescents. Third, the hypothesized relationship between media use and the endorsement of genderstereotypical sexual attitudes has been supported by empirical evidence (Ward, 2003) but has also been challenged with null findings (e.g., Ward & Friedman, 2006). In addition, some researchers have considered the theoretical basis for such sexual media effects to be inadequate (Collins et al., 2004), partly because those effects have commonly been explained by wide-ranging media effect theories, such as cultivation theory (e.g., Zurbriggen & Morgan, 2006). Although this criticism neglects to consider that studies of sexual media effects have sometimes developed more specific theories, such as the media practice model (Steele, 1999) or the heterosexual script theory (Kim et al., 2007), this view also indicates the need to test new theory-driven models. From this perspective, this study therefore explored whether objectification theory (Fredrickson & Roberts, 1997) and, more specifically, the three-step process of internalization, self-objectification and body surveillance may offer valuable insights into the effect of the media on sexual attitudes. To address our research goals, longitudinal data derived from adolescent boys and girls (N = 1,041) were used to test an integrative model on the relationships among the viewing of sexualizing sitcoms, internalization, self-objectification, body surveillance, and acceptance of the sexual double standard. This structural equation model showed that viewing sexualizing sitcoms (time 1) predicted internalization and self-objectification (time 2). In turn, internalization positively predicted body surveillance and self-objectification (all time 2). Self-objectification also predicted body surveillance (all time 2). Finally, internalization (time 2) positively predicted acceptance of the sexual double standard (time 3). The discussion focuses on the implications of these findings to explain the relationships among sexualizing media, self-objectification and the developing sexuality and body image of adolescents.
AN EXPERIMENTAL INVESTIGATION INTO THE EFFECT OF ONLINE SELF-PORTRAYAL ON WOMEN’S SELF-OBJECTIFICATION - DIAN DE VRIES AND JOCHEN PETER Girls and women are frequently treated as sexual objects, valued predominantly in terms of their physical and sexual attractiveness (Zurbriggen et al., 2007). At the same time, sexually objectifying content is pervasive in popular Western media (Zurbriggen et al., 2007). Evidence has accumulated that being sexually objectified or viewing sexual objectification of women in the media leads girls and women to “self-objectify” (Moradi & Huang, 2008). Self-objectification means that people view themselves “as a body” and focus on their physical appearance rather than on what they can do or how they feel (Fredrickson & Roberts, 1997), which can have serious negative consequences for wellbeing (Moradi & Huang, 2008).
25
26
ETMAAL 2013
BOOK OF ABSTRACTS
Existing research on the role of communication in self-objectification has been limited to face-to-face interactions and traditional media, notably magazines and television (e.g. Aubrey, 2006). This is surprising because, over the past years, the Internet has become a popular venue for social interaction and a medium in its own right (DiMaggio, Hargittai, Neuman, & Robinson, 2001). Scholars have observed a “rise of a new form of socialized communication: mass self-communication” (Castells, 2007, p. 248), which is characterized by users generating and distributing media content about themselves. A popular example of mass self-communication is keeping a personal profile on a social network site (SNS). The self-portrayals on SNS profiles are accessible to a large and unspecific audience (boyd, 2008). When girls and women portray themselves on SNS, their body and looks are frequently assessed and commented upon, and thus can expect they will be judged at least partly based on their physical appearance (Siibak, 2009). Research has shown that when women expect to be evaluated based on their appearance, this leads them to self-objectify (Calogero, 2004). Therefore, self-portrayal to an online audience may lead to self-objectification among women. It is the first aim of this study to test this. Existing objectification research is based on the assumption that sexually objectifying content, such as scantily dressed women in advertisements or on magazine covers, affects self-objectification in a cumulative, but largely independent way (e.g. Aubrey, 2006). However, when women portray themselves online, traditional sexually objectifying content may very well present a condition that boosts the potential effect of online self-portrayal on self-objectification, for example by making sexually objectifying ideas more cognitively accessible through priming (Roskos-Ewoldsen, RoskosEwoldsen, & Carpentier, 2009). In that case, exposure to traditional sexually objectifying content interacts with online self-portrayal in its effect on self-objectification. It is, therefore, the second goal of this study to test initially whether traditional sexually objectifying content moderates the effect of online self-portrayal on self-objectification. Method We conducted an online experiment with a two (priming stimuli: objectifying vs. neutral) by two (audience: online audience vs. no audience) between-subjects design among 221 women aged 18 to 25. In the objectifying priming stimuli condition participants were asked to evaluate two objectifying advertisements, whereas in the neutral condition participants were shown two non-objectifying advertisements. All participants created an online profile, which consisted of choosing an avatar and writing a self-description of five sentences. Participants in the online audience condition were told that their profile would be visible to others, whereas participants in the no audience condition were told that no-one would view their profile. To measure self-objectification, two independent coders assessed which percentage of the self-descriptions directly or indirectly focused on the own appearance. This method was similar to the Twenty Statements Test (TST) (Fredrickson, Roberts, Noll, Quinn, & Twenge, 1998), which is often used in experiments to measure self-objectification (Calogero, 2011). Interrater agreement was high (k = .91). Results The results showed an interaction between audience condition and priming stimuli condition on selfobjectification. More specifically, portraying the self to an online audience led to higher self26
27
ETMAAL 2013
BOOK OF ABSTRACTS
objectification in comparison with portraying the self privately, but only among women who had been primed with sexually objectifying media content. Conclusion The results suggest that women’s online self-portrayal, if combined with sexually objectifying stimuli, may induce self-objectification. Discussion Our findings suggest that, in order to obtain a comprehensive understanding of the role of communication in self-objectification in contemporary society, we need to 1) broaden our knowledge to include online interactions and media forms, 2) pay attention to traditional media content as a possible moderating factor in the process of self-objectification.
WOMEN’S REJECTION OF SEXUALLY EXPLICIT CONTENT: THE ROLE OF HYPERFEMININITY AND PROCESSING STYLE - JOHANNA M.F. VAN OOSTEN, JOCHEN PETER AND INGE BOOT In the past decades, evidence has emerged that women often show negative cognitive and affective responses to sexually explicit content (Laan, Everaerd, van Bellen, & Hanewald, 1994; Mosher & Maclan, 1994). Such rejection of sexually explicit content has previously been explained by the type of sexual content (Laan et al., 1994; Mosher & Maclan, 1994). Specifically, the higher prevalence of objectification and degradation of women in male-centered sexual content seems to result in a higher rejection by women of male-centered sexual content compared to female-centered sexual content. To date, however, it remains unclear which types of women are more likely to reject sexually explicit material. Moreover, scholars have suggested that the rejection of sexually explicit material results from women linking the content to their pre-existing sexual beliefs (Mosher & Maclan, 1994). But it is still understudied how precisely women come to link the sexual content to their sexual beliefs. The present study therefore had two goals: first, to investigate individual differences in sexual beliefs (i.e., hyperfemininity) in the effect of sexual content on rejection, and second, to explain this effect by type of processing women engage in (response- vs. stimulusfocused). Hyperfeminine women are generally more likely to accept sexual subordination and objectification of women, as well as male sexual dominance (McKelvie & Gold, 1994; Murnen & Byrne, 1991; Nowatzki & Morry, 2009). Such themes are prevalent in much of male-centered sexual content (Brosius, Weaver, & Staab, 1993; Cowan, Lee, Levy, & Snyder, 1988; Cowan & Dunn, 1994). Thus, when linking male-centered sexual content to one’s sexual beliefs, hyperfeminine women seem less likely to reject such content than non-hyperfeminine women. However, the link between pre-existing sexual beliefs and sexual content has only been assumed, but never been empirically demonstrated. To more thoroughly investigate whether it is indeed this linking of the specific sexual content to one’s beliefs that results in rejection of such content, we therefore focused on the specific way in which such content is processed. The model of cognitive processing of sexual stimuli (e.g., Dekker & Everaerd, 1988) posits that women are more likely to link the specific sexual content to their sexual beliefs when using stimulusfocused processing. Specifically, when stimulus-focused processing is used, associations with the situational context are activated in memory. As a result, when women engage in stimulus-focused 27
28
ETMAAL 2013
BOOK OF ABSTRACTS
processing of a male-centered sexual stimulus, cognitive associations (e.g., of sexual inequality and subordination of women) are activated that are more likely to be accepted by hyperfeminine women than by non-hyperfeminine women. We therefore expected an interaction effect between hyperfemininity and processing type: When using stimulus-focused processing, women will reject male-centered sexually explicit content (relative to female-centered sexual content), but only when they have low degrees of hyperfemininity. Further, if the type of processing indeed matters for the rejection of sexually explicit material, hyperfeminine and non-hyperfeminine women are unlikely to differ in their rejection of sexually explicit content when using response-focused processing: Response-focused processing activates associations in memory (e.g., sexual feelings) that are not expected to differ by degree of hyperfemininity. In an online experiment among young women aged 18-30 (N = 195), both the type of sexually explicit content (male- vs. female-centered), and processing style (response- vs. stimulus-focused) were manipulated. In addition, participants were divided into three groups (i.e., low, moderate, high) based on their degree of hyperfemininity. As expected, when using a stimulus-focused processing style, male-centered sexual content resulted in greater rejection of the content (i.e., counterarguing) compared to female-centered sexual content among women with low or moderate hyperfemininity, but not among women with high degrees of hyperfemininity. This effect of hyperfemininity was not found when participant used a response-focused processing style. The present study showed that individual differences in sexual beliefs such as hyperfemininity can indeed influence rejection to male-centered sexual content. Furthermore, we empirically demonstrated that this influence was the result of linking the content to pre-existing beliefs, by manipulating the way the content was processed. The current findings suggest that both individual differences and processing styles should be taken into account in order to improve our understanding of why women reject sexually explicit content.
MEDIA USE AND ADHD-RELATED BEHAVIORS IN CHILDREN AND ADOLESCENTS: A METAANALYSIS - SANNE NIKKELEN, PATTI M. VALKENBURG AND MARIETTE HUIZINGA Children’s media entertainment has changed considerably in the past decades. It has become more arousing, fast-paced, and violent, and has been targeting children at an ever younger age (e.g., Gentile & Anderson, 2003). Concurrently, the number of children diagnosed with attentiondeficit/hyperactivity disorder (ADHD) or who display ADHD-related behaviors have significantly increased (e.g., Newsom, Archer, Trumbetta, & Gottesman, 2003). ADHD is characterized by three main symptoms: inattentiveness, hyperactivity, and impulsivity (DSM-IV-TR; American Psychiatric Association, 2000). There has been a growing concern that the increasing occurrence of ADHD-related behaviors can be attributed to the changes in children’s media environment (e.g., Christakis, 2009). Although studies into the relationship between children’s media use and ADHD-related behaviors have accumulated, they have yielded inconsistent results. As a result, it is still impossible to conclude whether children’s media use and ADHD-related behaviors are related, and if so, under which conditions. To fill this gap 28
29
ETMAAL 2013
BOOK OF ABSTRACTS
in the current literature, we conducted a quantitative meta-analysis of studies investigating the relationship between screen media use and ADHD-related behaviors in children and adolescents up to the age of 18. Within this meta-analysis, we distinguished three outcome variables: inattentiveness, hyperactivity, impulsivity, and composite ADHD (i.e., combined measures of inattentiveness, hyperactivity, and impulsivity). The main goal of this study was to calculate an overall average effect size of the relationship between media use and these four outcomes. In addition, we examined three possible moderators: media characteristics (violent content, fast pacing, and overall time spent using media), media types (television, video games) and child characteristics (age, gender). Our sample included of 33 studies. Only few of the included studies used an outcome measure exclusively tapping inattentiveness, hyperactivity, or impulsivity. Therefore, meta-analyses could not be performed for the outcomes inattentiveness and hyperactivity. For impulsiveness, only an average effect could be calculated. Further analyses on media characteristics, media type, and child characteristics could only be performed for the outcome ADHD composite. The analyses revealed a significant, albeit small, positive relationship between media use and ADHD composite (r = .12) as well as between media use and impulsivity (r = .10). For the outcome ADHD composite, the average correlation coefficient of studies into television use (r = 0.11) was lower than that of videogame use (r = 0.16). The moderating effect of media type, however, was not significant. This might be caused by the low number of studies into videogame playing (N = 11) relative to the number of studies on television viewing (N = 33). Regarding media characteristics, the average correlation coefficient of studies into overall time spent using media was the same as that of studies into violent media use (r = 0.12). Too few studies examined the effect of fast-paced media to reliably calculate an average effect size. A meta-regression analysis with mean age as predictor did not indicate an effect of age on the effect size. A meta-regression analysis with the proportion of girls in the sample as predictor resulted in a negative effect on the strength of the correlation. This indicates that the relationship between media use and ADHD-related behaviors might be somewhat stronger for boys. We were not able to directly compare effect sizes between boys and girls because only one studies reported separate effect sizes. In summary, our results indicate a positive relationship between media use and ADHD-related behaviors, which might be stronger for videogame playing compared to television viewing and for boys compared to girls. These latter two results, however, are preliminary and need replication. The findings of this study indicate that future studies examining the relationship between media use and ADHD-related behaviors should not be focused on whether a relationship exists, but on possible factors of individual susceptibility that influence the strength of this relationship. Moreover, future studies should distinguish inattentiveness, hyperactivity, and impulsivity, because the effect of media use are likely to be different for these three outcomes.
29
30
ETMAAL 2013
BOOK OF ABSTRACTS
Session #4 Thursday 7 February 14:05 – 15:05 hrs. Chair: Martine van Selm Location: Aberdeen (M3-03) English/Dutch session
THE DARK SIDE OF ‘REALITY TV’: QUESTIONS OF ETHICS IN THE PRODUCTION OF ‘POPULAR FACTUAL ENTERTAINMENT’ TELEVISION - JELLE MAST To contend that ‘reality TV’ is morally vexed would strike few as an overblown or, for that matter, groundbreaking assertion. For this television phenomenon, it is fair to argue, has been a quite prolific topic of public debate, often invoking deep concerns over fundamental human values of dignity and respect for others, honesty or truthfulness, and integrity. Such public contestations tend to emerge, submerge, and re-emerge, in a repetitive movement, around individual, more or less extreme incidents. By the same token, the critical discourse is shaped by the contours of a ‘moral panic’ and by derogatory notions that (pre)conceive ‘reality TV’ as a ‘bad object’ – as ‘trash’, ‘voyeur’ or ‘humiliation’ television (see e.g., Calvert, 2004; Hill, 2007; Mills, 2004). So, an ‘incident-centered ethics’ (Evers, 2007) stands against the all-encompassing rhetoric of absolutist positions, which potentially illuminate structural factors and trends (political-economic, socio-cultural) but easily gloss over the complexities and nuances, subtle or more substantial, of ‘the field’. In this article we aim to strike a middle ground by developing a critical, comprehensive yet differentiating account of key moral issues in ‘reality TV’. Our focus is on the (potentially) harmful implications of, and the responsibilities that emerge from, the process of making ‘reality TV’, and the hybrid ‘reality’-show (Bondebjerg, 2002) variant in particular. This study is grounded in the concepts and insights provided in the literature on screen documentary, media and image ethics, set against the background of the recent popularization and hybridization of screen documentary. For this development complicates the straightforward transposition of either traditional ‘documentary-asjournalism’ ethical frameworks, or the ‘amoral efficiency’ of creative works in the fictional realm, thus somewhat disorienting our usual ‘moral compasses’ (Winston, 2000; Hill, 2007). Our analysis proceeds along three broad (not mutually exclusive) notions of ‘intrusion’, ‘humiliation’, and ‘misrepresentation’. In doing so, it integrates conceptualizations with empirical findings emerging from semi-structured interviews with 12 television professionals (mostly ‘creative’) and 25 participants (as well as a number of contracts). The interviewees had all been involved in formats with a border-crossing circulation, delivered by different television production companies to both public service and commercial stations. As such, this study moves beyond what could be inferred from the screened program to the extra-textual dimension, and contributes to the steadily growing area of empirical (‘reality’) television production studies. Mapping and schematizing pertinent moral issues related to the (harmful) treatment of participants in the production of ‘reality’-shows, illuminates the role of structural factors and the importance of differentiation, in kind and degree, regarding the moral qualities of the practices and products of 30
31
ETMAAL 2013
BOOK OF ABSTRACTS
individual television professionals and the larger entities they comprise. With reference to the ‘Potter Box’-model of moral reasoning and analysis (in Christians et al., 1998), our study thus informs the crucial steps of ‘empirical definition’ of the situation, and identification of (professional, aesthetic, moral) ‘values’ and ‘loyalties’, while arguing for the articulation of ‘ethical principles’ to guide decision-making. The case for a profound ethical (self-)critique of ‘reality TV’ is particularly relevant as, in many respects, the making of reality TV structurally accommodates certain morally dubious conditions and practices. For the political-economic structure of contemporary television imposes distinct (i.e. market-oriented, ratings- and budget-driven) imperatives and associated (commercial) values, while curtailing the ‘checks-and-balances’ built in ‘the system’ (through deregulation and fragmentation of the production process). More specific to ‘reality TV’, its hybrid nature and (relative) lack of an institutional framework comparable to journalism, further complicate ethical judgment (‘disorienting our moral compasses’). ‘Reality’-shows thrive on the increasingly porous boundaries between ‘public’ and ‘private’, and on pleasures derived from the performances of subjects ‘acting as themselves’ in contrived contexts, while tenuously balancing appeals to ‘creative license’ with claims to ‘reality’. Significantly, the ‘reality TV’-aesthetic also works against a conscientious compliance with the principle of ‘informed consent’ (besides commonsensical limits to its practical effectiveness). This further questions a purely legalistic, contractual approach to prior consent, and its deployment as a legitimate ethical touchstone in ‘consent defense’ discourses (Winston, 2000). The paper concludes, therefore, that it is pivotal to bring ethical considerations more squarely into the day-to-day calculations of the ‘reality TV’-profession, through the development of ‘moral accountability systems’ (Bertrand, 2000). The present study certainly contributes to this cause by stimulating a ‘moral pause’ (Gross et al., 1989) and by yielding theoretical and practical insights for the elaboration of an ethical framework for ‘reality TV’.
PRIME TIME MORALITY: A LONGITUDINAL STUDY OF MORALITY IN DUTCH PRIME-TIME FICTION SERENA DAALMANS, ELLEN HIJMANS AND FRED WESTER Over the time western cultures have moved from a gemeinschaftliche to a gesellschaftliche organization of society, as well as from modernity to postmodernity (Van Poecke, 2000, p.132). Society has changed through processes like individualization, rationalization and in-formalisation effectively transforming society’s focus from purely community-based thinking to incorporating the (absolute) importance of the individual (Aupers, Houtman & Van der Tak, 2003; Schnabel, 2004). This meant that society’s moral palette grew from a purely group focus community, to include elaborate professional and market-driven moralities and also grew to include moralities based on individual expressiveness, hedonism and pleasure (Tipton, 2002). When conceptualizing television as a reflection of society in its role as main storyteller and disseminator of accepted norms, behaviors and moral lessons (Gerbner, 2002), one can expect that these changes in morality will also be reflected within television programs over time. When studying the representation of morality on television, the connection with television fiction is easily made when considering that in the storytelling-function fiction programs form the most obvious vehicles for moral lessons about society’s conceptions of rele-vance and value. As Gerbner (1999/2002) pointed out, television fiction illuminates the hidden dynamics of the real world, how 31
32
ETMAAL 2013
BOOK OF ABSTRACTS
things work, how we should judge things in life and what value they have (good, bad). For some time now, television critics and scholars alike have remarked the growing narrative complexity of television fiction (Mittel, 2006; Johnson, 2005a, 2005b). Since the moral lesson of stories is embedded in the plot this development should lead to more morally complex plotlines. Furthermore, critics and scholars have also pointed to a dramatic increase in morally ambivalent lead characters in contemporary televi-sion fiction (Eden, Grizzard & Lewis, 2011; Krakowiak, 2008; Krakowiak and Oliver, 2009; Raney, 2004). Several case studies, focusing on this prevalence of the morally ambivalent main characters in television fiction, have illustrated variations of the type such as the anti-hero, the tragic villain and the gangster hero (Bokiniec, 2011; De Wijze, 2008; Donnelly, 2012; Harris, 2012; Imre, 2009; Keeton, 2002). We argue that morality in television fiction is connected to the plot as well as the main characters, and therefore expect that over time through the increase in narrative complexity in plotlines as well as the increase in ambiguous main characters the television fiction’s morality has grown increasingly complex. This study is the third part of an ongoing research-project studying morality and Dutch television, building upon the results from a previously conducted qualitative content analysis (Daalmans, Hijmans & Wester, 2012) and quantitative longitudinal analysis of socio-cultural changes represented on television (Emons, 2011) as well outcomes by other studies focusing on (aspects of) morality and television (e.g. Bilandzic, 2011; Costera-Meijer & Krijnen, 2007; Grabe, 2002; Selnow, 1986/1990; Sutherland & Siniawsky, 1982; Tavener, 2000). We now focused on morality in Dutch prime time fiction. We conducted a quantitative content analy-sis, with six coders, based on a longitudinal sample of Dutch prime-time fiction (1985-1995-2005) (N = 250). Our main research goal and question was to discern if and how fiction has changed over time and how this has affected the representation of morality on television. How has fiction changed over time and how does this affect television’s representation of morality? We divided this overall goal and question in three sub-questions focusing on three different aspects of fiction that when reviewing the literature on previous results: changes in life domains over time, changes in characters moral nature over time and changes in presence of different types of morality over time. RQ1: What life domains dominate in Dutch prime time fiction and has this changed over time? RQ2: Do main characters change over time regarding their moral nature? RQ3: Which moralities are represented in prime time fiction and does their presence change over time? Preliminary results point to changes in fiction regarding its representation of changes in soci-ety (e.g more female characters, less married characters, more young adults) and an increase in moral complexity due to changes on the level of the moral nature of characters and in terms of types of morality.
32
33
ETMAAL 2013
BOOK OF ABSTRACTS
BIEDT DEXTER ZIJN KIJKERS MOREEL VERMAAK? EEN GEVAL VAN “MORAL DISENGAGEMENT” MONIQUE TIMMERS, ED TAN, CLAIRE SEGIJN AND GUUS BARTHOLOME Theoretische rationale: De geneugten die drama te bieden heeft zijn veel bestudeerd. In de mediapsychologie was het Zilmann die met zijn dispositietheorie (DT) wees op het belang van sympathie voor de protagonist als voorwaarde voor plezier bij de toeschouwer. De DT gaat er bovendien vanuit dat sympathie op haar beurt bepaald wordt door morele attitudes. Toeschouwers van drama zijn onophoudelijk bezig personages en hun handelen op moraal te controleren. Aardig is wie goed is of doet, onaardig wie slecht doet of is. De traditie van misdaaddrama, het genre dat in dit onderzoek is bestudeerd, wil dat de protagonist-politieman moreel goed is, en diens tegenstander, de antagonist-dader, slecht. De inspecteur of rechercheur is dus sympathiek, de misdadiger onsympathiek. Sinds de jaren negentig komen in misdaaddrama’s steeds vaker en nadrukkelijker protagonisten en antagonisten voor die moreel niet zwart-wit zijn getekend. Bekende misdaadtelevisiedrama’s als Breaking Bad en Dexter zijn interessante voorbeelden. In Dexter, de serie die in dit onderzoek is bestudeerd, is de hoofdpersoon werkzaam voor de justitie en tevens seriemoordenaar. Misdaaddrama waarin moreel ambigue personages (Krakowiak & Oliver, 2004) een hoofdrol vervullen stelt de DT voor een probleem. Hoe kunnen we genieten van doen en laten van slechte personages? Recent heeft Raney (2004) betoogd dat sympathische disposities bij het kijken naar dramapersonages onmiddellijk post vatten. Morele oordelen volgen pas in tweede instantie. Meestal zijn ze in lijn met bestaande sympathieën en versterken ze die. Maar het kan ook zijn dat een reeds sympathiek bevonden personage blijk geeft van een amorele instelling. Om te blijven genieten van personage en verhaal moet deze evidentie als het ware onschadelijk worden gemaakt. Kijkers zullen volgens Raney overgaan tot Moral Disengagement (MDis), loskoppeling van het morele oordeel over het personage van de vaste eigen morele opvattingen. Het morele oordeel over het personage is een gelegenheidsoordeel, aangepast aan de sympathie, en morele overtuigingen moeten daarvoor wijken. Naar onze mening is er nog een andere verklaring voor het behoud van plezier bij het kijken naar een immorele protagonist. Entertainmentproducties zoals misdaaddrama kan men zien als speelse oefeningen in empathie, sympathie en moraal. Fictie en entertainment zijn speciaal ingericht voor het doen van zulke oefeningen. Wij veronderstellen dat Dexter als entertainment-misdaadkomedie wordt opgevat, en meer in het bijzonder dat morele loskoppeling (MDis) door de kijkers leuk wordt gevonden om aan mee te doen ( althans in fantasie of fictie). Vandaar de hypothese die wij in deze studie testen, namelijk dat Mdis een factor is in de receptie die los van sympathiserende dispositie t.o.v. de hoofdpersoon plezier en waardering voor de serie zal voorspellen. Methode: 209 deelnemers aan de online survey keken naar Dexter, Season 1 Episode 5, Love American Style en vulden daarna een vragenlijst in bestaande uit 7-punts likert items. Metingen: Moral Disengagement: omschrijvingen van Raney (2004) aangevuld met interviews om te komen tot een lijst van items die worden geacht Mdis te meten, bijv. “Bij het kijken naar de serie liet ik allerlei morele overwegingen varen”, “Dexter vermoordt alleen schuldige mensen, dus het is rechtvaardig wat Dexter doet”, en “Het is wel erg wat er in de serie gebeurt, maar het is niet echt, het blijft verzonnen.” Plezier werd gemeten door de vraag hoe leuk de episode was en naar een rapportcijfer (1-10). Sympathie en empathie voor Dexter werd door drie items gemeten. Voorts werd gevraagd naar attitudes t.a.v. eigenrichting, gêne ervaren tijdens het kijken, en demografische 33
34
ETMAAL 2013
BOOK OF ABSTRACTS
respondentkenmerken. Resultaten: Meervoudige regressie-analyses brachten aan het licht dat 1) Sympathie/ empathie significant bijdraagt aan Plezier; 2) MDis significant bijdraagt aan Sympathie/ empathie, en 3) MDis bijdraagt aan Plezier als gecorrigeerd is voor de bijdrage van Sympathie/ empathie. Attitude t.a.v. eigenrichting droeg niet bij aan Plezier, terwijl Gegeneerdheid er afbreuk aan deed. Demografische kenmerken zijn in de analyses constant gehouden. De uit de gegevensanalyses getrokken conclusie is dat plezier in het kijken naar Dexter berust op sympathie, in overeenstemming met de DT van Zillmann. Er is bovendien steun voor Raney’s veronderstelling dat het loslaten van je morele principes sympathie met een immoreel handelen personage helpt de sympathie te bewaren, in zoverre dat MDis sympathieverhogend werkt. Voorts blijkt dat het loslaten van morele principes op zichzelf leuk wordt gevonden, het verhoogt het plezier. In de discussie gaan we in op details van moral disengagement bij Dexter. Een inhoudsanalyse levert stof voor het vermoeden dat Dexter twee kanten heeft: Diabolische psychopaat en Engel der Wrake. Moral disengagement zoals door ons gemeten helpt Dexter te zien als Engel der Wrake.
THE RELATIONSHIP BETWEEN FICTION VIEWING AND SUPPORT FOR THE POLICE: THE IMPORTANCE OF PROCEDURAL FAIRNESS BELIEFS - ASTRID DIRIKX AND JAN VAN DEN BULCK Research consistently finds support for a procedural justice perspective on public support for the police. According to this perspective overall support for the police is mostly a reaction to beliefs about how the police deal with people: whether they treat people fairly and respectfully, recognize citizen rights and care about people’s concerns (Bradford & Jackson, 2010; Reisig & Lloyd, 2009). Communication scholars, and cultivation researchers in particular, have long been interested in the effects of television’s portrayal of crime and law enforcement on people’s conceptions of reality (e.g. Gerbner & Gross, 1976; Surette, 2007). Very few studies have, however, looked at the relationship between media use and beliefs about the procedural fairness of the police. The way police officers deal with community residents is a staple of much entertainment programming and recent content analyses of popular police series show that procedural justice-relevant content is regularly featured (Dirikx et al., 2012; Eschholz et al., 2004). The current study aims to examine the relationship between exposure to television fiction and perceptions of the procedural fairness of the police. In addition, the study will explore procedural fairness judgments as a mediating factor between television exposure and instrumental evaluations of the police. Research indicates that procedural fairness beliefs are directly related to support for the police, but also indirectly by influencing beliefs about police effectiveness and beliefs about how equally the police distribute their services across communities (distributive fairness). Van den Bulck (2012) argues that cultivation researchers cannot automatically assume a similar impact of native and foreign-produced television programs on viewer’s conceptions of their native culture. In the current study we choose to focus on the process of primary cultivation: Is watching Flemish police shows related to how people think about the police in Flanders? Based on content analyses results it was expected that exposure to Flemish police shows on the public channel would be positively related to beliefs about procedural fairness (H1) and that exposure to police shows on the commercial channels would be negatively related to procedural fairness perceptions (H2). 34
35
ETMAAL 2013
BOOK OF ABSTRACTS
Procedural fairness beliefs were hypothesized to be a positive predictor of support for the police (H3). The same was expected for perceptions about the performance and distributive fairness of the police (H4). A final hypothesis predicted procedural justice judgments to be positively related to perceptions about police performance and distributive fairness (H5). Surveys were administered to a sample of 1968 secondary school adolescents. Adolescence is the most critical period for legal socialization (Piquero et al., 2005). Of the respondents, 51.4% were boys. The mean age was 16 (SD = 1.83). 46.2% were in ASO, 26.4% in TSO and 27.4% in BSO. 86.6% of the respondents thought other people primarily saw them as Belgian. 13.4% were convinced others mostly saw them as of non-Belgian descent. The hypotheses were tested through the use of structural equation modeling. The hypothetical model fitted the sample data very well (χ² = 1260.27, p = .00; CFI = .95; RMSEA = .04). The model explained 19% of the variance in perceived procedural fairness, 46% of performance beliefs, 40% of distributive justice judgments and 61% of respondents’ overall attitude toward the police. Exposure to police shows on the public channel was positively related to perceptions of procedural fairness (γ = .12, p < .05). The opposite was found for exposure to shows on the commercial channels (γ = -.13, p < .05). Procedural fairness perceptions were a strong direct predictor of respondents’ global attitude toward the police (γ = .43, p < .05) and were also positively related to performance judgments (γ = .62, p < .05) and distributive justice evaluations (γ = .57, p < .05). Bootstrap analyses showed that procedural fairness beliefs fully mediated the relationships between police show exposure and instrumental beliefs about the police. The indirect effects of watching police shows on the public channel on performance perceptions (γ = .08, p < .05) and on distributive justice evaluations (γ = .07, p < .05) were significant. The same was found for the indirect relationships between exposure to series on the commercial channels and views on the performance (γ = -.08, p < .05) and distributive fairness of the police (γ = -.08, p < .05). The study shows procedural fairness is an important concept to integrate in research on media effects on support for the police. In addition, it is argued that procedural justice offers an interesting opportunity for cultivation researchers to study the way television cultivates shared conceptions of society’s norms and values.
35
36
ETMAAL 2013
BOOK OF ABSTRACTS
Session #5 Thursday 7 February 15:15 – 16:15 hrs. Chair: Gertjan Willems Location: Aberdeen (M3-03) English/Dutch session
NATIONAL CULTURAL REPERTOIRES OF EVALUATION IN A GLOBAL AGE: FILM DISCOURSE IN FRANCE, THE NETHERLANDS, THE UNITED KINGDOM, AND THE UNITED STATES - ANNEMARIE KERSTEN The expanded international exchange of culture products caused Western cultural fields to resemble each other. Hollywood movies prevail on import markets around the world, inducing homogenized film fields dominated by blockbusters. However, whereas global audiences to a large degree consume the same movies, they do not necessarily make sense of them in the same manner. Cultural surroundings may still differentiate the socially constructed national cultural repertoires of evaluation. This article examines the sustenance of such repertoires through the analysis of film criticism in France, the Netherlands, the United Kingdom, and United States. Combined qualitative and quantitative methods demonstrate that while critics in all countries employ the same discourse components, evaluative repertoires vary in composition and style. Western countries command distinguishing features that particularize their manners of meaning making despite the ubiquity of globalizing trends.
DE (IDEOLOGISCHE) REPRESENTATIES VAN WETENSCHAP IN DE RAMPENFILM - LAURENS VAN DER STEEN AND PIETER MAESEELE Rampencinema wordt door wetenschapscommunicatieve benaderingen gebruikelijk geanalyseerd in termen van ‘effectiviteit’. Welke impact heeft de rampenfilm op bepaalde indicatoren van bepaalde gewenste attitudes bij de kijkers van de rampencinema? Bijvoorbeeld: Is de kijker van een rampenfilm na het zien van de film meer geïnteresseerd in de gevolgen van klimaatverandering? Deze ‘traditionele’ benadering gaat echter voorbij aan de ideologische kenmerken van rampenfilms, die zich voornamelijk bevinden in de representatie van wetenschap. Deze paper wil aantonen op welke manier deze films bijdragen aan maatschappelijk hegemonische representaties waarin bepaalde ‘ongewenste’ wetenschappelijke praktijken worden gemaskeerd en voorgesteld als inherent wenselijk. Dit gebeurt aan de hand van een diepgaande kwalitatieve inhoudsanalyse op drie voorbeelden uit het genre: The Day After Tomorrow, Contagion en The Rise of Planet of Apes. Traditionele wetenschapscommunicatie heeft tot op vandaag weinig aandacht voor de analyse van de rampenfilms. Door het toenemende belang van maatschappelijke discussies over klimaatverandering heeft de traditionele wetenschapscommunicatie haar roeping gevonden als de noodzakelijke wetenschap die de communicatie tussen wetenschap en de maatschappij moet stroomlijnen. De latente assumptie is dat de overdracht van informatie noodzakelijk bijdraagt aan de ‘correcte’ oplossing van het probleem klimaatverandering. Wat wij willen beargumenteren is dat de 36
37
ETMAAL 2013
BOOK OF ABSTRACTS
opkomst van klimaatverandering niet louter vraagt om op de traditionele manier aandacht te schenken aan ‘vergeten’ mediavormen – zoals de rampenfilm –, maar dat het roept om een nieuwe benadering van wetenschapscommunicatie. Daarvoor baseren we ons onder andere op het werk van de Duitse socioloog Beck die stelt dat onze maatschappij steeds meer de vorm van een laatmoderne maatschappij aanneemt. Klimaatverandering, GGO’s, kernenergie, etc. zijn in het denken van Beck zogenaamde moderniseringrisico’s, die een ongewenst gevolg zijn van de rationele onderwerping van de leefomgeving in de moderniteit. Relevant voor communicatiewetenschappers is dat deze risico’s discursieve openheid kennen in oorzaken, werking, betekenis en gevolgen. Men kan klimaatverandering niet via een klassieke juridische handeling-schade-vergoeding-logica toeschrijven aan een specifieke maatschappelijke groepering. Er ontstaat dus een noodzakelijk maatschappelijk conflict over wie de kosten zal dragen voor een risico als klimaatverandering. Uit de resultaten van de analyse van de geselecteerde films blijkt een zeer specifieke ideologische representatie van wetenschap en haar diëgetische omgeving naar voren te komen. Het meest basale onderscheid is gebaseerd op een eenvoudige dichotomie tussen wetenschappers en nietwetenschappers, met specifieke stereotypen voor zowel politici, media en zakenlui. We beargumenteren dat wetenschappers in onze films in principe terugkerend twee rollen aannemen, en dat zij als enige in staat geacht worden om de maatschappij te redden van de ramp alsook van de maatschappij zelf. Deze bevindingen zullen zowel op een statische als een dynamisch-narratologische wijze worden geanalyseerd. Opvallend is dat ondanks de grote verschillen in narratief en thematiek we kunnen aantonen dat deze films een grote homogeniteit vertonen wat betreft de ideologische kenmerken van de representaties van wetenschap. Deze blootgelegde ‘ideology of science’ blijkt vergelijkbaar met die uit onderzoek naar andere mediavormen. Verder wordt besproken welke potentieel negatieve gevolgen deze ideologische representaties kunnen hebben op democratisch vlak, bijvoorbeeld wat betreft laatmoderne risico’s zoals de teelt van GGO’s. We besluiten met een bevraging van mogelijke filmische alternatieven voor deze ideologische representaties. Hieruit zullen we enkele ideeën ontwikkelen over het nut van het analyseren van de ideologie van wetenschap in rampenfilms (en de bredere populaire cultuur), alsook suggesties doen voor verder onderzoek in rampenfilms.
DE AFBEELDING VAN STATUSVERSCHILLEN IN SEKSE EN KLEUR IN NEDERLANDSE NON-FICTIE PROGRAMMA’S - SJOERD PENNEKAMP, CLAIRE SEGIJN, GUUS BARTHOLOMÉ AND MONIQUE TIMMERS Televisie is het meest gebruikte medium in Nederland. In 2011 keek de gemiddelde Nederlander 3 uur en 11 minuten per dag naar de televisie (SKO Jaarrapport, 2011). Er is geen ander medium waaraan dagelijks zoveel tijd wordt besteed (SPOT 2012). Van alle tijd die men aan televisie besteed bestaat ruim 70 procent uit non-fictie programma’s. Omdat mensen een belangrijk deel van hun informatie tot zich nemen via de televisie, beïnvloed dit medium in sterke mate hoe zij de wereld om zich heen waarnemen (Gerbner & Gross, 1976). Omroepen zelf realiseren zich ook dat zij een belangrijke rol vervullen in de beeldvorming. Niet alleen het evenwichtig representeren van de maatschappij is een speerpunt van de Publieke Omroep. Ook pluriformiteit - het laten horen van verschillende meningen uit de maatschappij – geldt als speerpunt.
37
38
ETMAAL 2013
BOOK OF ABSTRACTS
De status die aan verschillende groepen wordt toegedicht is een belangrijke kenmerk dat door afbeeldingen in de media wordt beïnvloed. Doordat sommige groepen veel vaker dan anderen in rollen met een hogere status worden afgebeeld, kan het beeld ontstaan dat deze groepen ook een hogere maatschappelijke status hebben. Zo is volgens Gerber (2009) de lagere status die in mediauitingen worden toegekend aan vrouwen verantwoordelijk voor de perceptie dat vrouwen en mannen verschillende karaktereigenschappen hebben. Daarbij komt dat personen met een hogere status ook vaker om hun mening worden gevraagd. Dit leidt er toe dat de mening van de lage status roepen veel minder wordt gehoord. Het onevenwichtig afspiegelen van verschillende groepen kan dus leiden tot een vertekend beeld van deze groepen in de samenleving. Stereotypes worden mede gevormd door wat mensen zien en horen in de media (Babad, Birnbaum, & Benne, 1983). Televisie in het bijzonder kan een rol spelen bij het vormen van seksegerelateerde stereotypen (Signorielli & Bacue, 1999). Deze stereotypen beïnvloeden niet alleen de reacties op deze groep door andere groepen. Ook leden van de lage statusgroep zelf kunnen het idee krijgen dat zij niet geschikt zijn voor functies met een hoge status. Eerder onderzoek naar de afbeelding van minderheden (bijv. Mastro & Greenberg, 2000; Mastro & Stern, 2003) is met name verricht in het buitenland. Deze studies onderzoeken representatie bovendien vaak in de context van fictie of reclames. In Nederland zijn veel minder studies gedaan naar de afbeelding van verschillende groepen in televisieprogramma’s (echter zie Emons, Wester & Scheepers 2010; Koeman, Peeters & D’Haenens, 2007). In deze studie onderzoeken we daarom hoe de status van vrouwen en kleur wordt gerepresenteerd in Nederlandse non-fictie programma’s en hoe deze verschilt van respectievelijk de status van mannen en witte mensen. We doen dit door te kijken naar de rollen die deze verschillende groepen in programma’s vervullen. Voor dit onderzoek werd in de maand maart van 2010 van de zenders van de Publieke Omroep, RTL en SBS een week lang bijgehouden in welke rollen verschillende groepen werden afgebeeld. Dit werd gedaan voor alle Nederlandse non-fictie programma’s op deze zenders, in totaal werden in 99 uitzendingen 1443 personen gecodeerd. Van de personen die sprekend in beeld kwamen werd het geslacht en de kleur vastgelegd. Ook werd bijgehouden in welke rol de personen werden afgebeeld. Daarnaast werd gecodeerd welke functie de personen bekleden en of zij deskundig waren ten aanzien van het onderwerp waarover werd gesproken. Ook de gespreksonderwerpen zelf werden gecodeerd. De resultaten laten ten eerste zien dat er veel minder vrouwen dan mannen te zien zijn in Nederlandse non-fictie programma’s. Ook kleur komt minder voor in de gemeten programma’s dan in de maatschappij. Vrouwen en gekleurde mensen worden niet alleen minder frequent afgebeeld, als zij worden afgebeeld is dit ook vaak in functies met een lagere status. Zo worden vrouwen minder vaak afgebeeld als interview subject of als sprekers voor groepen. Net als gekleurde mensen zijn vrouwen wel vaker te zien als kandidaten in spelletjes. Personen in hoogopgeleide functies zijn ook minder vaak vrouw of gekleurd, terwijl studenten en scholieren juist vaker vrouw of gekleurd. Vrouwen en gekleurde mensen zijn ook veel vaker te zien als vox populi dan als deskundige. De onderwerpen waar zij over spreken zijn – met name in het geval van vrouwen – minder zwaar. Vrouwen spreken veel meer over consumentenzaken en relaties, terwijl onder mannen onderwerpen als politiek en economie juist vaker voorkomen.
38
39
ETMAAL 2013
BOOK OF ABSTRACTS
In de discussie gaan we in op de mogelijke gevolgen van de statusverschillen op stereotypen over deze groepen. Daarbij belichten we de rol die televisie kan spelen in het bevorderen van een gebalanceerde afbeelding van de status van verschillende groepen.
FILM HISTORY BETWEEN CULTURE AND ECONOMICS: ADMISSION FIGURES AT THE CAPITOLE CINEMA (1953-1971) - KHAËL VELDERS, LIESBETH VAN DE VIJVER AND DANIËL BILTEREYST Traditionally film studies are focused on the stylistic, narrative and cultural history of cinema. Even with the acknowledgement that the movies were cultural yet commercial artifacts traded on a variety of film markets, contemporary film studies favor a socio-cultural film history as a niche. An inherent distrust of social science empirical research methods and little willingness to broaden the scope of dominant textual analysis and medium specific perspectives still force other kinds of film historiography in the shadows. Research on the economical factors of the local film business remains open for research. Our contribution examines local film culture based on the analysis of actual attendance figures from the major first-run cinema of Ghent, the Capitole, between 1953 and 1971. It is a historiographic research on unique archival material captured in a database containing data on 801 unique films. Based on cultural economics, the analysis explores issues like taste-patterns, financial policy, cultural entrepreneurship and programming strategies in a pivotal period of crisis in the film market. The Capitole was a major first-run theatre in Ghent from 1932 to the introduction of the multiplex in the city’s centre in 1982. Capitole was a commercial cinema, void of pillarization or vertical integration and run by the Lummerzheim family. Over the years it became a perfect example of going the movies translated in an elaborate price differentiation, a full-sized staff, glamorous atmosphere and expansive services. The paperwork of public limited liability company Cinex-Sofexim, responsible for the management of Capitole and six other centre cinemas in Ghent, were archived in the national archives containing balance books, programming documents, correspondence, reports on meetings, photos, blueprints, etcetera. The archive can therefore be subject to many historic questions. This study focused on admission figures provided for the period 1953 to 1971. The data was imported in a database identifying the original film titles combined with data on the screening, the ticket price, the gross revenues, the taxes, the admission numbers and the net income. A first analysis describes the programming strategies of Capitole based on preferences in genre of the films, their country of origin, actors, year of production, production company or other noteworthy characteristics of the programmed films. A second analysis studies the financial health of the first-run cinema in a crucial period of crisis in cinema culture. Based on these analyses, Capitole is characterized as an exclusive first-run cinema thriving on sheer reputation and keen entrepreneurship. Quantifying Capitole’s programming enabled us to gain insight into this cinema’s commercial profile and longitudinal trends in program changes. Unsurprisingly, results show a share of 40.5% in American programming. However, a rather remarkable measurement is a share of 23.1% in FrenchItalian coproductions; rising near the early sixties and peaking in 1964, only to decline again. Accompanying these data is the observation that 77.2% of these coproductions are French language movies; signifying a certain preference of Capitole’s taste-community. Secondly, our analysis asserts Capitole’s position as a first-run cinema by revealing a high share of premiere movies, which nevertheless declines steadily towards the early seventies as the movie theatre enters financially 39
40
ETMAAL 2013
BOOK OF ABSTRACTS
difficult times. Another exercise we have made deconstructs programming in relation to movie genre. Results show that war movies, comedies, adventure films, drama’s and historical movies accounted for 53.3% of all movie screenings – with the comedy genre by itself construes 22.6% of the movie screening total. Overall, Capitole seemed to lack a strong genre profile. The further analysis shifts focus from profiling programming to financial operations. Gross income – adjusted for inflation and interpreted as the result of the number of sold tickets, advertising revenue, government support and consumption sales – is seen declining steadily, but levelled around the mid-sixties to then grow slightly again. This observation can be aligned with the historical crisis in cinema culture by pointing out Capitole’s ticketing strategies: the average ticket price, adjusted for inflation, almost doubles over a span of 15 years to compensate the substantial decrease of attendance, which drops over 50% from 1953 to 1971. The methods of analysis used in this study offer a concrete analysis of the inner economical working of the cinema business. It furthermore provides information on a specific taste-community and enables us to map patterns of cultural consumption.
Session #6 Friday 8 February 09:00 – 10:00 hrs. Chair: Jeroen Jansz Location: Santander (M2-11) English/Dutch session
A LONGITUDINAL STUDY INTO THE EFFECTS OF ADOLESCENTS’ SOCIAL NETWORK SITE USE ON THEIR APPEARANCE INVESTMENT AND DESIRE TO UNDERGO COSMETIC SURGERY - DIAN DE VRIES, JOCHEN PETER, PETER NIKKEN AND HANNEKE DE GRAAF The popularity of social network sites (SNS) among adolescents has stimulated considerable research on the role that SNS play for important aspects of adolescents’ development (e.g. Valkenburg, Peter, & Schouten, 2006). However, research on adolescents’ body image is scarce. The use of SNS is likely to influence adolescents’ body image because evaluating one’s own and others’ physical attractiveness is an important part of SNS use (Siibak, 2009). According to objectification theory (Fredrickson & Roberts, 1997), the significance of physical attractiveness on SNS may lead more frequent SNS users to become more invested in their appearance. However, it is also conceivable that individuals who are more invested in their experience use SNS more often. Expanding on a recent correlational study on the issue (Vandenbosch & Eggermont, 2012), it was the first aim of the current study to explore, with a longitudinal design, the causal direction of the relationship between SNS use and appearance investment among adolescents. Increased investment in appearance predicts desire to undergo costly and risky cosmetic surgery among adult women (von Soest, Kvalem, Skolleborg, & Roald, 2006), and these findings may also translate to adolescents. Cosmetic surgery among adolescents, in particular, is controversial because their body and body image are still developing (Sarwer, Infield, & Crerand, 2009). If SNS use enhances investment in physical appearance, and if this appearance investment in turn increases adolescents’ 40
41
ETMAAL 2013
BOOK OF ABSTRACTS
desire to undergo cosmetic surgery, SNS use may indirectly affect this desire. It was the second goal of the present study to investigate this indirect effect. Research into appearance investment and cosmetic surgery has typically focused on girls and women. However, adolescent boys also experience pressures to look physically attractive on SNS, albeit to a lesser extent (Siibak, 2009). Therefore, SNS use may also lead to appearance investment and, indirectly, to desire to undergo cosmetic surgery among adolescent boys. However, little research to date has looked into the effects of SNS on adolescent boys’ body image and it is unknown if the relationships between SNS use, appearance investment, and desire to undergo cosmetic surgery are more apparent, or exclusive, to girls. The third aim of the current study was to investigate this. Method The current study is a secondary analysis of a two wave panel survey conducted in 2008 and 2009. The surveys, which investigated adolescent’s (sexual) media use, sexual attitudes and behaviors, and body image, was administered twice, six months apart. For the current study, we were specifically interested in the measures regarding frequency of SNS use, appearance investment and desire to undergo cosmetic surgery. These measures were completed in full at both time points by 604 Dutch adolescents (age 11-18). Results We tested our hypotheses using structural equation modeling, following recommendations by Cole and Maxwell (2003). Results of both parametric tests and bootstrapping procedures showed that adolescents’ SNS use and appearance investment reinforced each other. Furthermore, increased SNS use indirectly lead to desire to undergo cosmetic surgery through appearance investment. Multiplegroup analyses showed that these relationships were not moderated by gender. Discussion Our findings suggest that when unraveling the relationships between communication and adolescents’ body image, 1) the role of online communication should not be neglected, 2) we should consider media effects as a mutually reinforcing dynamic rather than as one-directional, conform the media effects approach suggested by Slater (2007), and 3) we should not neglect the experiences of boys and men.
CAN THE EXPERIENCE OF PLAYING A SHOOTER GAME BE AS DIFFERENT AS CHALK AND CHEESE? AN INVESTIGATION OF HOW PLAYERS CONSTRUE THEIR PLAYING STYLE - WANNES RIBBENS AND STEVEN MALLIET Research on the relationship between exposure to violent video games and aggression-related outcomes has often focused on the direct impact of violent game play (e.g. Carnagey, et al., 2007). Some researchers point out that this may impede a more profound understanding of how and why media use affects individuals (Peter & Valkenburg, 2010, p.390). Because video games have been characterized as a medium in which players actively contribute to the content or message (e.g. Simkins & Steinkuehler, 2008) it is important that we understand how games are played in order to explain their impact on violence or aggression (Weber, et al., 2009). Nonetheless the question of how 41
42
ETMAAL 2013
BOOK OF ABSTRACTS
players construe their playing style has remained relatively under-examined, especially with regard to violent games (Clarke & Duimering, 2006; Weber, et al., 2009). From a game design perspective a number of player archetypes have been proposed, of which the taxonomy by Bartle (1996) remains the most widely used, despite it having been criticized for not being empirically grounded (Yee, 2006). From a game theory point of view, structural analyses have addressed the affordances and limitations imposed by the rules of a game (e.g. Juul, 2005), but have rarely considered how the player interacts with these rules (Malaby, 2007). Quantitative content analyses (e.g. Smith, et al, 2003) offer an account of the narrative ingredients of a game but downplay the multitude of experiences that arise during game play (Schmierbach, 2009). Studies that categorize players based upon their motivations (e.g. Yee, 2006) have only sporadically accounted for player-game interactions and have mainly restricted themselves to online multiplayer games. Finally, ethnographic work (e.g. Steinkuehler, 2005) provides extensive theoretical but limited empirical grounds because usually only one game and a few players are under scrutiny . In this study we use methods from Human-Computer Interaction and the social sciences to further our understanding of how players of shooting games construe their playing style. We aim to examine 1) whether players differ in how they play a shooting game 2) whether these differences can be observed between players across different shooting games, and 3) whether these differences can be found within players of the same shooting game. 26 male students of a methodological seminar course on digital games played a shooting game for at least 12 hours. They were required to fill in a qualitative diary structured around an analysis scheme after every 45 minutes of game play. In order to capture the dynamic of game play ten respondents participated in a retrospective think aloud protocol. All participants, with the exception of two, participated in a FGI in which they were encouraged to reflect on their playing style through comparison with other gamers. A comprehensive discussion about the advantages and disadvantages of using these methods to study player experiences can be found in Ribbens & Poels (2009). An inductive analysis of the data according to the principles of Wester & Peters (2004) revealed several themes that were used by the participants to describe their gaming behaviour. In general, these could be reduced to four continuums or axes of play: Gameplay-driven vs narrative-driven; exploration vs efficiency; reactive play vs cognitive play; and risk-taking vs prudence. The results revealed large difference across players of different games as well as within players of the same game. Classifying players based on how they scored on all axes of play proved very difficult and could only be described, in imitation of Steinkuehler (2005), as complicated messiness. Nevertheless, by using a reduction matrix and by considering the axis ‘reactive play vs cognitive play’ and the axis ‘risktaking vs prudence’ as primary concepts, six player archetypes could be discerned: Run and gun players, achievers, clear thinking attackers, clear thinking defenders, realism seekers, and reactive players. The results of this study demonstrate that players differ in how they experience and process the content of shooting games, and provide a first step towards operationalizing these differences within a quantitative framework. Interestingly, these differences were robust between players of different games but also between players of the same game, which suggest that players have a considerable degree of power to tweak the game play experience. The findings were partially in line with research 42
43
ETMAAL 2013
BOOK OF ABSTRACTS
on playing styles in other game genres (Bartle, 1996; Steinkuehler, 2005) and strong similarities were observed with research on game involvement (Calleja, 2007).
PROFIEL VAN COMPULSIEVE SNS-GEBRUIKERS EN GAMERS IN BELGIË SURVEYONDERZOEK BIJ VOLWASSENEN IN VLAANDEREN EN WALLONIË - ROZANE DE COCK, JOLIEN VANGEEL, ANNABELLE KLEIN, PASCAL MINOTTE, OMAR ROSAS AND GERT-JAN MEERKERK Opzet studie en methode In België zijn tot nog toe geen grootschalige betrouwbare gegevens beschikbaar over de omvang van compulsief computer en internetgebruik en het specifieke profiel (sociodemografische en psychosociale kenmerken etc) van mensen die kampen met compulsief SNS- en gamegebruik. Voor dit onderzoek werden in opdracht van Federaal Wetenschapsbeleid (Belspo) 1000 volwassenen (18+) uit België bevraagd over hun internet- en computergebruik met specifieke aandacht voor sociale netwerksites (SNS) en games (online en offline). De steekproef is representatief voor de bevolking naar geslacht, regio (Vlaanderen, Wallonië en Brussel), leeftijd en opleiding. Om valide uitspraken te kunnen doen over een voldoende grote groep compulsieve versus non-compulsieve gebruikers van zowel SNS als games, werden voor de uiteindelijke dataset respondenten gefilterd die minstens 16 uur per week online zijn. Literatuur ivm compulsief internetgebruik Onder onderzoekers bestaat geen consensus over de terminologie die problemen met excessief internetten (Tisseron, 2008 ; Valleur, 2006) aanduiden. Sommigen spreken over “verslaving” (Griffiths, 1998; Young, 1998), anderen kiezen voor “compulsief internetgebruik”(Greenfield, 1999; Meerkerk, 2007), “problematisch” (Caplan, 2002) of “pathologisch internetten” (Davis, 2001). In navolging van Meerkerk en van den Eijnden & Vermulst (2006) kiezen wij voor de term compulsief internetgebruik. Frenk & Dar (2000, pp. 20-21) definiëren compulsie als “specific kinds of bad habits, consisting of dysfunctional, purposeful and repetitive behavior routines. […] It is usually attached to behaviors that carry short-term pleasure or relief of stress, but negative long-term consequences”. Een extra argument hiervoor is het bijhorende meetinstrument (CIUS, Compulsive Internet Use Scale), bestaande uit 14 items, dat in een ruime populatie gevalideerd is (Meerkerk, 2007). De items in deze schaal meten deelkenmerken van CIU: het verlies van controle, mentale en gedragsmatige preoccupatie, ontwenningsverschijnselen, stemmingsschommelingen en conflict. Voor het meten van compulsief gamegebruik gebruiken we de Videogame Addiction Test (VAT) (van Rooij et.al., 2012) en voor compulsief SNS-gebruik pasten we de BFAS (Bergen Facebook Addiction scale) aan tot een CSS (Compulsive Social Networking Site Scale, gevalideerd in 2012, eigen onderzoek). Algemene beschrijving steekproef De steekproef bestaat voor 50% uit vrouwen en voor 50% uit mannen, 56.6% is Vlaming, 33.7% Waal en 9.7% Brusselaar. De gemiddelde leeftijd bedraagt 43. Meer dan de helft van de respondenten werkt, 9% is werkloos, 15% is met pensioen en 9.30% studeert nog. Profiel compulsieve versus non-compulsieve SNS-gebruiker 79,2% van de respondenten heeft minstens één SNS-profiel. In volgorde van populariteit komen de volgende SNS aan bod: Facebook, Twitter, LinkedIn, Netlog en MySpace. Google+, Hyves, Pinterest en MSN zijn veel minder vaak gebruikt in België. De gemiddelde leeftijd binnen de groep compulsieve 43
44
ETMAAL 2013
BOOK OF ABSTRACTS
SNS-gebruikers (CS) is 35 en de meerderheid is vrouw (61.5%). De tijd die zij spenderen aan SNS tijdens een school- of werkdag (gemiddeld 2 uur en 40 minuten) ligt zoals verwacht significant hoger dan bij niet-compulsieve gebruikers (gemiddeld 1 uur en 32 minuten) (t(70.204) = -3.645, p < .05). Dit tijdsgebruik neemt toe tijdens een weekenddag of in de vakantie (3 uur en 7 minuten voor CS en 1 uur en 51 minuten voor niet-compulsieven. Beide groepen verschillen niet wat extraversie en vindingrijkheid betreft. CS scoren gemiddeld lager op emotionele stabiliteit. Dit is ook het geval voor aardigheid en nauwgezetheid. CS voelen zich eenzamer en hebben meer last van depressieve gevoelens. Tot slot scoren ze met een gemiddelde van 31.26 lager op eigenwaarde dan nietcompulsieve gebruikers met een score van 36 (t(790) = 6.805, p < .05). Bij de nog school lopende volwassenen (N=143) is er een verschil te merken in attitude tegenover school: CS scoren gemiddeld lager (5.1/10) in vergelijking met niet-compulsieve SNS-ers (7.3/10) wanneer ze gevraagd worden aan te geven hoe graag ze school lopen. Bij de werkende groep geven CS aan meer van zichzelf te vinden dat ze tijdens de werkuren te veel tijd spenderen aan internetten voor privé-zaken (t(571) = -3.06, p < .01) en zelfs al op de vingers getikt zijn door de baas. Profiel compulsieve versus non-compulsieve gamer De compulsieve gamers (CG) zijn gemiddeld 34 jaar en 60.3% is man. Ze besteden op een school- of werkdag gemiddeld 2 uur en 35 minuten aan offline gamen. Voor een niet-compulsieve gamer (NCG) ligt dit op 1 uur en 27 minuten. Ook op een dag in het weekend en de vakantie wordt significant meer tijd besteed aan offline gamen door CG: 3 uur en 56 minuten versus 2 uur en 8 minuten. Hetzelfde gaat op voor online gamen. Wanneer CG vergeleken worden met NCG naar het soort games, blijken ze vaker actie/FPS-games (χ2 = 9.284, p < .05, df = 1), vechtgames en MMORPG’s te spelen. CG verschillen van NCG op de persoonlijkheidseigenschappen aardigheid en nauwgezetheid. Op het vlak van extraversie, vindingrijkheid, emotionele stabiliteit en eenzaamheid worden er geen verschillen gevonden.. CG blijken meer depressieve gevoelens en minder eigenwaarde te hebben.
HET VERBAND TUSSEN GEWELDDADIGE GAMES EN DESENSITISATIE: DE ROL VAN GEOBSERVEERD REALISME EN SPEELSTIJLEN - STEVEN MALLIET AND WANNES RIBBENS In deze paper wordt de relatie tussen het spelen van 'shooting games' en desensitisatie tegenover geweld onderzocht, en wordt bestudeerd in hoeverre de concepten geobserveerd realisme en speelstijl een modererende of mediërende rol spelen in deze relatie. Desensitisatie wordt gedefinieerd als een verminderde cognitieve of emotionele respons tegenover geweld (bijvoorbeeld Funk, et al., 2004). Waar vele onderzoekers een direct verband vaststelden tussen blootstelling aan gewelddadige games en desensitisatie (bijvoorbeeld Arriaga, et al., 2011; Carnagey & Anderson, 2005), vonden anderen geen of slechts een gedeeltelijke directe samenhang (bijvoorbeeld Colwell & Kato, 2003; Bowen & Spaniol, 2010). Een mogelijke verklaring bevindt zich in het feit dat vaak verschillende operationaliseringen van desensitisatie worden gebruikt. Om hierin duidelijkheid te scheppen neemt deze studie zowel cognitieve als culturele en emotionele attitudes tegenover geweld in rekening als maatstaven van desensitisatie. Een andere verklaring wordt aangereikt door recent onderzoek dat suggereert dat het effect van gewelddadige videogames zich niet op een directe wijze manifesteert, maar wordt gemedieerd en/of 44
45
ETMAAL 2013
BOOK OF ABSTRACTS
gemodereerd door variabelen die verband houden met de speler-spel relatie. Deze studie onderzoekt de tussenliggende rol van twee dergelijke factoren: geobserveerd realisme (Ribbens & Malliet, 2010) en snelheid van spelen (Adachi & Willoughby, 2011). Bijkomend werden, in navolging van Ferguson en collega's (o.a. Ferguson, et al., 2008; Ferguson, 2011), een brede waaier aan persoonlijkheids- en contextuele kenmerken in rekening gebracht omdat deze mogelijk een betere voorspeller zijn van agressieve uitkomsten dan mediablootstelling. Een vragenlijst werd opgesteld die de volgende concepten in rekening bracht: Onafhankelijke variabelen: - Blootstelling aan geweld in 'shooting games' - Persoonlijkheidskenmerken: anti-sociale persoonlijkheidsschaal (Paternoster & Mazerolle, 1994), depressieve symptomenschaal (Cole, et al.., 2004). - Sociale context: opleidingsniveau, steun van anderen (Ferguson, et al., in pers), geweld/persoonlijke slachtofferschaal (Gonzales, et al., 1995), delinquente 'peers' (Paternoster & Mazerolle, 1994), huiselijk geweld (Sherin, et al., 1998). Tussenliggende variabelen: - Geobserveerd realisme: simulationeel realisme (Ribbens, 2012, in pers). - Snelheid van spelen Afhankelijke variabelen: - Emotionele desensitisatie: empathie (Funk, et al., 2008). - Cognitieve desensitisatie: attitudes tegenover geweld schaal en cultuur van geweld schaal (Funk, et al., 1999). - Gewelddadig gedrag: delinquentie schaal van Ferguson, et al. (in pers). De vragenlijst werd afgenomen bij 1846 jongens met een gemiddelde leeftijd van 15,7 jaar die werden gerekruteerd via een cluster sampling van scholen. Hierbij werd een lichte oververtegenwoordiging van studenten in het TSO-onderwijs en een lichte ondervertegenwoordiging van studenten uit het BSO-onderwijs vastgesteld. Initiële analyses van de modererende (volgens de methode van Aguinis, 2004) of mediërende rol (volgens de bootstrap methode van Preacher & Hayes, 2008) gaven aan dat geobserveerd realisme en speelsnelheid niet optreden als moderators van het verband tussen blootstelling aan gewelddadige games en desensitisatie, maar dat ze wel een mediërende rol kunnen spelen. Deze mediërende rol werd verder onderzocht door middel van twee 'structural equation' modellen die werden opgesteld, andermaal gebruik makende van de bootstrap methode van Preacher & Hayes (2008), maar met inclusie van de variabelen gerelateerd aan persoonlijkheid en sociale context. Het model met betrekking tot de cognitieve component van desensitisering heeft een aanvaardbare fit (χ²/df=2.37, CFI=.92, RMSEA=.033 (CI90: .031-.035), en geeft aan dat blootstelling aan 45
46
ETMAAL 2013
BOOK OF ABSTRACTS
gewelddadige games zwak gerelateerd is aan attitudes tegenover geweld (β=.060, B=.001, SE=.001, p=.044), maar niet aan gewelddadige cultuur. De verwachte relaties tussen geobserveerd realisme en gewelddadige cultuur (β=.220, B=.139, SE=.019, p<.000), tussen geobserveerd realisme en attitudes tegenover geweld (β=.169, B=.148, SE=.027, p<.000) en tussen speelsnelheid en attitudes tegenover geweld (β=.135, B=.113, SE=.027, p<.000) worden ondersteund, terwijl het verband tussen speelsnelheid en gewelddadige cultuur niet significant blijkt. Analyses van de indirecte samenhang tussen blootstelling aan gewelddadige games en de afhankelijke variabelen, en van de directe samenhang tussen blootstelling aan gewelddadige games en de tussenliggende variabelen, geven aan dat geobserveerd realisme en speelsnelheid een mediërende rol spelen in het verband met attitudes tegenover geweld. In het geval van de afhankelijke variabele 'gewelddadige cultuur' treedt enkel geobserveerd realisme op als een mediator. Opleidingsniveau, antisociale persoonlijkheidskenmerken, delinquente ‘peers’ en slachtoffer van geweld zijn significant geassocieerd met cognitieve desensitisatie voor geweld. Een gelijkaardig model werd opgesteld met betrekking tot de emotionele component van desensitisatie. In dit geval treedt enkel speelsnelheid op als mediator, en wordt voor geobserveerd realisme geen significante samenhang vastgesteld. Opleidingsniveau, antisociale persoonlijkheidskenmerken, sociale steun en depressie zijn significant geassocieerd met emotionele desensitisatie voor geweld. Als conclusie worden deze resultaten besproken in het licht van de bestaande literatuur, en wordt een pleidooi gemaakt voor een speler-gerelateerde aanpak van game effect onderzoek, waarbij de speler-spel relatie een prominente rol krijgt toebedeeld.
Session #7 Friday 8 February 10:15 – 11:15 hrs. Chair: Margot van der Groot Location: Santander (M2-11) English session
THE RELATIONSHIP BETWEEN ON-LINE AND OFF-LINE FEAR OF CRIME: A MEDIATED MODEL KATHLEEN CUSTERS AND JAN VAN DEN BULCK Fear of crime has received considerable attention in media effects research. At least two types of fear have been studied. The distinction between both types is based on whether or not TV viewing coincides with experiencing the emotion. The first type will be referred to as off-line fear of crime in this study. It refers to feelings of fear regarding crime which people experience in the real world, when they are not exposed to media messages. In media effects research this type of fear is considered to be a “major focus of the cultivation theory” (Morgan & Shanahan, 2010, p. 341). The central tenet of this theory is that heavy viewing may lead to a distorted perception of the world as a mean and dangerous place which in turn may affect fear in the real world (Gerbner & Gross, 1976; Gerbner, Gross, Morgan, & Signorielli, 1980). As early as 1976, Gerbner and Gross (1976) stated that “people who watch a lot of television are more fearful in the real world” (p. 41). While media effects 46
47
ETMAAL 2013
BOOK OF ABSTRACTS
scholars examine whether TV viewing contributes to real world fear, this type of fear obviously occurs in people who have no access to television as well. A second type of fear which has been examined in media effects research will be referred to as online fear of crime. On-line fear occurs when violent TV images trigger fear responses while viewing. The present study examines on-line fear of crime as a “self-referent emotion” (Dunlop, Wakefield, & Kashima, 2008, p. 55). This classification of emotions refers to emotions triggered by viewers’ thoughts about their own life that are activated by the exposed message. The current study empirically tests a model in which experiencing on-line fear of crime predicts real world judgments (such as perceived risk, perceived control and perceived seriousness of crime victimization), which in turn predict levels of off-line fear of crime. It will draw on research on how emotion affects judgment, most notably the affect-heuristic (Finucane, Alhakami, Slovic, & Johnson, 2000; Slovic, Finucane, Peters, & MacGregor, 2004, 2007), the risk-as-feelings hypothesis (Loewenstein, Weber, Hsee & Welch, 2001), and the appraisal tendency framework (ATF) (Lerner & Keltner, 2000, 2001). These theories support the notion that emotion can effect judgment. Data were collected in a random sample of 711 adults by means of a standardized self-administered questionnaire. A two step randomization model was used for selecting participants. The sample consisted of 43% men. The average age in the sample was 44.9 years (SD = 17.4) ranging from 18 to 93. Of the respondents 21.1% had finished 9th grade or less, 33.4% only finished 12th grade; 24.4% finished two or three years of college and 21.1% finished 4 or more years of college or university. The results of both an exploratory and confirmatory factor analysis showed that on-line and off-line fear of crime are two distinct concepts. In addition, the goodness-of-fit indices of a structural equation model produced an acceptable fit to the data (RMSEA = .065; CFI = .95; NFI = .93; CMIN/df = 4.031). On-line fear of crime was positively related to perceived risk of crime victimization (β = .31, p < .0001) and negatively related to perceived ability to cope with crime (β = -.09, p < .01). These perceptions, in turn, were significant predictors of off-line fear of crime. Crime risk perception was positively correlated with off-line fear (β = .46, p < .0001) whereas perceived ability to cope with crime was negatively related to off-line fear of crime (β = -.09, p < .05). The model explained 27%, of the variance of off-line fear of crime. The significance of the indirect effects was tested with the bootstrapping method. The results from a bootstrapping procedure (500 samples, ML bootstrap, Monte Carlo procedure, CI 95%) showed that the standardized indirect effect of on-line fear of crime on off-line fear of crime was significant (.152, SE = .026) at p <.01. Results are discussed with respect to the application of the Appraisal Tendency Framework in media effects research. With regard to the cultivation hypothesis, the current study suggests that apart from the traditional pathway going from the amount of television viewing to first- or second-order judgments, there is at least one other pathway: from experiencing emotions during television exposure to judgments. Thus, while the traditional view on cultivation theory sees television exposure as the cultivator of perceptions, the current study illustrates that the emotions experienced while viewing may cultivate judgments as well.
47
48
ETMAAL 2013
BOOK OF ABSTRACTS
CONSTRUCTING AND VALIDATING A SCALE TO ASSESS MORAL JUDGMENT OF ANTISOCIAL MEDIA CONTENT AND MEDIA PREFERENCES AMONG YOUTH - XANTHE S. PLAISIER AND ELLY A. KONIJN Given the popularity among youth of media content portraying antisocial and norm-crossing behaviors such as unhealthy and risky behaviors (e.g., reckless driving, drugs abuse, bullying; Brown & Witherspoon, 2002; Funk & Buchman, 1996), there is a need for research on understanding adolescents’ preferences for this kind of antisocial media content. Many studies have pointed at the negative consequences of exposure to such antisocial behaviors in the media (Anderson et al., 2010; Strasburger, 2009; Strasburger, Jordan, & Donnerstein, 2010). However, the question why or when adolescents would prefer this kind of media content has been largely understudied, particularly in systematic and experimental research designs. Furthermore, hardly any research is available on how youngsters evaluate such adverse media content, particularly in view of their moral evaluations. Consequently, no appropriate measurement device is currently available to assess adolescents’ media preferences and moral judgment of (antisocial) media content in a systematic way. Therefore, we constructed a measurement tool to assess adolescents’ preferences for and moral judgment of antisocial media content, titled the Media Morals and Youth Questionnaire (MMaYQue), which did a good job in a previous study (Authors). However, in constructing a new measurement device, it is of utmost importance to validate the scale in view of future usage. Therefore, the aim of the present study is to validate the MMaYQue in multiple samples in light of its psychometric qualities, construct, discriminant, and criterion validities. METHOD The core of the MMaYQue contains 15 descriptions of popular YouTube-clips, edited and presented as if they were original YouTube clips. Nine clips contain antisocial behaviors based on four domains of risk-taking behavior: violence/aggression, sexual harassment/sexual self-exposure, substance abuse/binge drinking, and reckless driving. A sample item is: “Two boys sexually harass a girl in the schoolyard”. To make the scale more balanced, six items describing neutral and prosocial behaviors were added as filler items. To measure media preference, the participant rates each clip description on a 5-point rating-scale to indicate to which extent s/he would like to watch the clip (1=not at all, 5=very much). To measure moral judgment, the participant evaluates each clip on two items to the extent s/he evaluates the behavior in the clip as acceptable and “OK” (1=not acceptable/OK, 5=very acceptable/OK). PARTICIPANTS AND MEASURES The MMaYQue was tested in two independent adolescent samples (NA=666, M=14.75, SD=1.58 and NB=792). Participants completed a paper-and-pencil questionnaire including the MMaYQue, general media usage (Krahé & Möller, 2011), preference for mature media content (Boxer et al., 2009), the Buss-Perry Aggression Questionnaire (Buss & Perry, 1992), short sensation seeking (Nije Bijvank, Konijn, Bushman, & Roelofsma, 2009), and the Index for Empathy for Children and Adolescents (Bryant, 1982). RESULTS The factor structure of the MMaYQue was tested using LISREL8.80 (Jöreskog & Sörbom, 1993). All 48
49
ETMAAL 2013
BOOK OF ABSTRACTS
models tested were based on the correlation matrix of the observed variables. Parameters were estimated using maximum likelihood estimation. Missing data were removed using listwise deletion. Assessing the fit of the models is based on whether the model fits better than rival models (using a chi-square difference test) and whether the model provided a good absolute fit. A model with an oblique two-factor structure (i.e., media preference and moral judgment) with correlated error terms, fitted best with the data in both samples of adolescents (sample A: RMSEA=.07; NFI=.95; CFI=96; GFI=.92; AGFI=.89; sample B: RMSEA=.08; NFI=.95; CFI=96; GFI=.90; AGFI=.86). Because the media preference and moral judgment questions are based on the same clips, it is reasonable to presume correlated error terms between the variables. A means and covariance structures approach (MACS; Sörbom, 1974) showed that the MMaYQue measures the same two constructs in various groups (i.e., in boys and girls, and in heavy vs. light media users; ∆CFI<.01). In addition, results showed good discriminant validity: Boys and heavy media users more strongly prefer, and are morally more tolerable towards, antisocial media content than girls and low media users. Finally, the MMaYQue evidences good criterion validity through its expected relationships with trait aggressiveness, sensation seeking, empathy, and preferences for mature media content. CONCLUSIONS The results of validating the MMaYQue in various samples on various criteria, shows a valid and reliable scale to assess adolescents’ media preferences and moral judgment of antisocial media content. The MMaYQue contains two related subscales: 1) preference for antisocial media content and 2) moral judgment of antisocial media content. The MMaYQue seems very promising for systematic research on adolescents’ media use and effects, as it is a generic measure that can be applied in both experimental, cross-sectional, and field studies.
THE ROLE OF MEDIA WITH ANTISOCIAL CONTENT IN CYBERBULLYING AMONG ADOLESCENTS ANOUK H. DEN HAMER, ELLY A. KONIJN AND MICHA G. KEIJER Cyberbullying behavior (in which bullying behavior takes place through electronic media, e.g. internet and mobile phones) has attracted considerable research attention over the last few years. Research showed that this type of bullying behavior is especially prevalent under adolescents and further demonstrated that between 20% and 40% of adolescents have been victim of cyberbullying behavior (Tokunaga, 2010). Between 30% and 65% of these victims do not tell their parents about the victimization(Sumter & Valkenburg, 2011) and it has been shown that victims of cyberbullying behavior have higher suicidal ideation rates than non-victims (Li, 2007; Hinduja & Patchin, 2010). These numbers underpin the emergency to determine the factors that stimulate cyberbullying behavior. The current research integrates three line of thoughts in various studies in cyberbullying in explaining how cyberbullying may develop in adolescents. First, research shows that victims of face-to-face bullying often become bullies themselves in cyberspace (Ybarra, Mitchell, 2004; Smith, Mahdavi, Carvalho, Fischer, Russel & Tippett, 2008). A second line of research shows that face-to-face victimization is strongly related to feelings of anger (Patchin & Hinduja, 2006; Beran & Li, 2007; Didden, Scholte, Korzilius, de Moor, Vermeulen, O’Reilly, Lang & Lancioni, 2009; Ortega, Elipe, MoraMerchán, Calmaestra & Vega, 2009). And a third area of research shows that adolescents seem to 49
50
ETMAAL 2013
BOOK OF ABSTRACTS
use violent media to cope with their anger (Arnett, 1996; Flammer & Schaffner, 2003; Olson, Kutner, Warner, Almerigi, Baer & Nicholi, 2007; Olson, Kutner & Warner, 2008; Bushman & Withaker, 2010). However, these lines of research have thus far not been integrated into a coherent chain of events. Furthermore, the role of media exposure in cyberbullying has hardly been investigated while adolescents are avid media users. Thus far, the limited results available suggest that exposure to violent media is related to cyberbullying behavior (Calvete, Orue, Estévez, Villardón & Padilla, 2010; Fanti, Demetriou & Hawa, 2012). The current study expands the investigation of the role of media in cyberbullying in focusing on a broader array of media fare. That is, we investigated the role of exposure to media with antisocial content more in general than violent content alone (i.e., including risky behavior, substance abuse, gossiping, etcetera), which also includes antisocial behavior more typical for girls. In this developmental stage, adolescents show an increased preference for media with antisocial content (Roe, 1995; Funk & Buchmann, 1996). Combining these lines of thought, the current study proposes and tests a cyclic process model in cyberbullying arguing that being face-toface bullied instigates feelings of anger and frustration, which draws the adolescent to media with antisocial content, subsequently reinforcing him/her to perform acts of cyberbullying behavior. Participants and Design Participants in a cross-sectional design were 892 adolescents (57% boys) aged 11-18 years (Mage =13.73, SDage=1.36). They participated voluntarily and were recruited from 16 different schools throughout the country. Variables measured in the cross-sectional survey included face-to-face victimization (Schwartz, Farver, Change & Lee-Shin, 2002), anger and frustration (Patchin & Hinduja, 2010), cyberbullying behavior (Calvete et al., 2010) and exposure to media with antisocial content (Authors, in press). The latter was measured with the newly developed Content-based Media Exposure Scale (CobME). Main Results and Discussion Results of the cross-sectional study revealed a first step in empirical support for the assumed cyclic process. We applied structural equation modeling (SEM) to test the theoretical model. The fit indices showed a good fit between the empirical data and the theoretical model good: χ2(688)=1920.13; CFI=.90, RMSEA=.05; p=ns. A significant relation between face-to-face victimization and feelings of anger and frustration was shown (β = .24, p < .001). In addition, the model indicated that exposure to media with antisocial content significantly mediated between anger/frustration and cyberbullying behavior (sum of indirect effect β=.14, p<.001). In all, the results of the cross-sectional study provided first evidence of a Cyclic Process Model in the chain of being bullied or victimized and the resulting anger/frustration leading them to exposure to media with antisocial content, which was then significantly related to cyberbullying behavior among adolescents. Thus, the results suggest that, mediated by feelings of anger and frustration, face-toface victimization draws one to media with antisocial content in stimulating their own bullying behavior in cyberspace. Longitudinal research is needed to further test the assumed Cyclic Process Model.
50
51
ETMAAL 2013
BOOK OF ABSTRACTS
DOUBLE DOSE: HOW VIOLENT MEDIA EXPOSURE AND FAMILY CONFLICT INTERACT TO PREDICT ADOLESCENTS' AGGRESSION - KARIN FIKKERS, JESSICA TAYLOR PIOTROWSKI, HELEN VOSSEN AND PATTI M. VALKENBURG The effect of violent media exposure on aggressive behavior has been investigated for decades, with most results pointing to small to moderate effect sizes (Anderson et al., 2010). Despite the current debate about whether these effect sizes are theoretically and practically meaningful, these relatively small effects sizes are neither strange nor unexpected. Aggressive behavior is a complex behavior caused by multiple factors (Browne & Hamilton-Giachritsis, 2005), of which violent media exposure is only one. Research investigating the effects of media violence in concert with other predictors of aggression is needed in order to better understand the relative contributions of media violence on aggressive behavior (Browne & Hamilton-Giachritsis, 2005). Doing so can provide important information about those individuals who are most susceptible to the effects of media violence. Like media violence, family context has been identified as an important factor in the development of aggression, particularly during childhood and early adolescence (e.g., Ribeaud & Eisner, 2010). Children growing up in high conflict families may learn that aggressive behavior is an acceptable way of interacting with others and consequently become more aggressive (Farver, Xu, Eppe, Fernandez, & Schwartz, 2005). Although violent media exposure and family conflict have both been investigated as predictors of aggression (e.g., Gentile, Coyne, & Walsh, 2011; Tanaka, Raishevich, & Scarpa, 2010), this has taken place in separate disciplines. Communication researchers have focused on the effects of violent media exposure on aggression, while ignoring effects of family conflict, while the reverse is true for family researchers (Vandewater, Lee, & Shim, 2005). Yet, both violent media exposure and family conflict provide early adolescents with contexts for learning aggressive behavior (Jordan, 2004). It is therefore interesting and relevant to look at both the individual and interactive effects of violent media exposure and family conflict on adolescents’ aggressive behavior. Using data from a two-wave longitudinal study among 499 Dutch adolescents aged 10 to 14 (47.9% girls; age at Time 1: M = 11.87 years, SD = 1.46), we investigated how media violence and family conflict predict aggression. Violent media exposure was measured by asking adolescents to report the number of days per week and number of hours per day they view violent television and play violent games. Family conflict was measured using five items about the frequency of conflict in a household (e.g., “How often do family members hit each other?”). Aggressive behavior was measured using six items that asked how often in the past six months an adolescent had done the following things to another adolescent: “call names”, “push in a rough way”, “kick or hit”, “threaten to beat up”, “fought with” or “tripped on purpose”. Because 86% of the study sample consisted of sibling pairs, data was analyzed with a multilevel modeling approach. This analysis takes into account the nested nature of the data (that is, siblings nested within households), which results in more correct estimates of the coefficients than ordinary least squares regression. In the analysis, we predicted Time 2 aggression from Time 1 violent media exposure and family conflict, while controlling for Time 1 aggression. Results showed a significant two-way interaction between violent media exposure and family conflict. The effect of violent media exposure on adolescents’ aggressive behavior was only 51
52
ETMAAL 2013
BOOK OF ABSTRACTS
significant for very low and very high levels of family conflict. A marginally significant (p = .056) threeway interaction with gender indicates that this pattern is different for boys and girls. For boys, the pattern was as expected: higher violent media exposure and higher family conflict resulted in higher predicted values for aggressive behavior. The pattern was the exact reverse for girls. Girls with higher levels of family conflict and higher levels of violent media scored lower on aggressive behavior. This unexpected finding could be explained by the fact that girls, generally, have a tendency to channel problem behavior into internalizing problems (e.g., depression, anxiety) rather than externalizing problems (e.g., aggression, McKelvey et al., 2011). These results highlight the importance of evaluating the effects of media violence within its larger context. The findings from this study suggest that family conflict can mitigate or exacerbate the effect of media violence, and that this pattern differs by gender. Future research on the differential effects of media violence is needed. Further, given the gender results in this study, investigations on more subtle aggressive behavior (e.g., relational aggression) as compared to external aggressive behavior would be an important next step.
Session #8 Friday 8 February 11:45 – 12:45 hrs. Chair: Tonny Krijnen Location: Santander (M2-11) English/Dutch session
THE RELATIONSHIP BETWEEN PARENTAL WORK DEMANDS AND CHILDREN’S TELEVISION TIME INE BEYENS AND STEVEN EGGERMONT While there is some evidence for positive outcomes of television exposure, several studies among young children have found television to be associated with sleep disorders (e.g., Thompson & Christakis, 2005), obesity ( e.g., Dennison & Edmunds, 2008), and attention problems (e.g., Zimmerman & Christakis, 2007), among others. In this respect, the American Academy of Pediatrics (AAP) recommends to limit screen media exposure among young children (AAP, 2009; 2011). Nonetheless, young children spend substantial amounts of their time with television (Rideout & Hamel, 2006). Because young children’s time with television is to a large extent controlled by their parents, it is important to understand what causes parents to have their children watch television. Research has revealed that parents, and working parents in particular, often feel unable to devote substantial time and energy to their children or to actively involve with their children, which makes it difficult to reduce children’s viewing time (Evans, Jordan, & Horner, 2011). An explanation for this may be found in time conflict theory (Eby, Casper, Lockwood, Bordeaux, & Brinley, 2005), which states that people are urged to divide their time and energy between work and family. Along these lines, it has been shown that more demanding work characteristics decrease the time parents spend with their child (e.g., Roeters, Van Der Lippe, & Kluwer, 2010). For instance, Brown, McBride, Bost and Shin (2011) showed that parents who worked longer hours played less with their child. However, the impact of parental work demands on the amount of children’s television viewing time remains largely unexamined. Therefore, the current study investigates the impact of parental working hours 52
53
ETMAAL 2013
BOOK OF ABSTRACTS
on the time children spend watching television. More specifically, the aim of this study is to contribute to the existing literature by testing hypotheses derived from time conflict theory (Eby et al., 2005) in an integrative model. The first hypothesis will test whether higher parental work demands, as indicated by parents’ working hours, result in parental distress and, subsequently, parenting time pressure, which in turn discourages parents to participate in shared activities with their children. Because scholars have argued that how parents engage with their children is important (Huston & Rosenkrantz Aronson, 2005), the second hypothesis will test whether the impact of work demands is stronger for shared play than for shared television viewing. Finally, the question arises as to how these shared activities relate to children’s television time. In this regard, we will test two hypotheses predicting that shared play and shared television viewing are related to children’s television viewing time. Data from 855 parents of children aged six months to six years (M = 3.04, SD = 1.713) are reported. An integrative model was constructed and structural equation modeling was performed to assess the hypothesized relationships. All analyses were performed using AMOS. Results showed that high working hours make parents feel more time pressured in their parenting, β = .13, B = .01, SE = .00. However, contrary to our expectations, this relationship was not mediated by parental distress, because parents’ working hours were not a significant predictor of parental distress, β = -.01, B = -.00, SE = .00. Subsequently, parenting time pressure discourages parents to play together with their child and to watch television together with their child (β = -.37, B = -.69, SE = .09 for play; β = -.11, B = -.28, SE = .11 for television viewing). The critical ratio for differences showed that the impact of working hours, mediated by parenting time pressure, was significantly stronger on shared play than on shared television viewing (crfd = 2.997). Finally, the findings revealed that less parent-child shared play increases children’s television time, β = -.17, B = -1.08, SE = .24, and that less parent-child television viewing decreases children’s viewing time, β = .21, B = 1.00, SE = .18. This study emphasizes the role of parents’ work demands as an important contributor to children’s television time. The findings of this study have several important implications. First, being less involved in watching television with a child, may well decrease the child’s total television time, it may also imply that the child watches more television unsupervised. Second, the finding that less parentchild play increases children’s television time, implies that the more developmentally appropriate parent-child play is displaced by time spent watching television, which is a less cognitively stimulating activity. The latter finding should especially warrant us, because parents’ work demands appear to be most influential on parent-child play.
HOW AND WHY PARENTS GUIDE THE MEDIA USE OF BABY’S, TODDLERS, AND KINDERGARTNERS; AN EXPLORATORY STUDY - PETER NIKKEN Children use audio-visual media at even younger ages (Rideout and Hamel 2006; Rideout e.a. 2010; Rideout 2011; Gutnick, e.a. 2011; Ofcom 2012). They do so because there are more media productions for younger audiences (Nikken, 2003; Valkenburg, 2008), more media devices available (De Haan & Pijpers, 2010), and because contemporary media devices are more easy to use. In a sample of over 1.500 parents, this study explores how and why parents guide young children’s 53
54
ETMAAL 2013
BOOK OF ABSTRACTS
media use in three different age groups (baby’s, toddlers and kindergartners), giving specific attention to parents’ views of media in the life of their children, the competences of their children to use digital media, and the activities their children perform on digital media devices. Parents’ views on media for children Parents are the most important agent for younger children when it comes to mediating the child’s media use, e.g. by posing restrictions on time and content, giving instructions, or by co-using the media with the child together. This mediation is guided by expectations about positive and negative effects from the media on their children (e.g. Valkenburg et al. 1999; Nikken & Jansz 2006; Livingstone & Helsper 2008; Sonck, Nikken & De Haan forthcoming). These studies have also indicated that parental mediation and the views on media effects are similar for different types of media. Moreover, research also showed that the views on effects and the amount of mediation vary with the parent’s demographics, media use etc. Research on parental mediation so far has had little attention for other implicit ideals or explicit ideas that parents can have regarding their up growing child, e.g. whether the child is competent enough to use the media or whether the media fulfil certain needs for the child, such as being entertained or giving a break. Also, parental mediation studies to date have not yet indicated why parents let their young children use media devices in their bedrooms. Parental mediation and child development The extent that parents apply mediation on the child’s media use is also associated with the types of media content the child is using, such as educational websites or social media applications (Cranmer e.a., 2006; Küter-Luks et al. 2011; Lee and Chae 2007; Sonck et al. forthcoming). Furthermore, mediation activities are also strongly related to the child’s age; some types of mediation are applied more often with older children, whereas other mediation styles are applied more often with younger children. So far, however, mediation studies have not yet looked into the relationship between children’s use of specific types of media content and their preferences and competences to use those media. In other words, there is a void in mediation studies with regard to aspects of child development and how parents adapt their views on media in the life of their developing child and their mediation styles. This study specifically aims to explore mediation by parents with younger children in relation to what children do with media and how skilled they are in using the media, and in relation to the views of the parents on media in the live of their children. Results Parents discerned four different roles of media in the child’s life: educational, harmful, too difficult yet, and providing a moment of rest. Regression analyses showed that young children in particular didn’t use media or had no media devices in their bedroom when parents saw the media as too difficult for their children. Views on risks or benefits of the media were not related. Regression analyses also showed that parents who perceived educational benefits of media co-used the media with their children in all age groups, and among baby’s restricted their media use. Among toddlers and kindergartners restrictions were more often used by parents when they saw risks of the 54
55
ETMAAL 2013
BOOK OF ABSTRACTS
media, and for kindergartners also if they felt the media are too difficult. Restrictions were also more often applied by parents with toddlers who were more competent in basic computer skills and parents with kindergartners who were more skilled in surfing. Moreover, parents also applied restrictive mediation when their toddlers and kindergartners started to play educational games and when kindergartners played shooting games. Overall conclusion Parental views on the role of media in the life of their children in general were more strongly associated with the parent’s mediation styles than the child’s media activities and competences. Apparently, parental ideals for raising children are more decisive for the parent’s behavior than the child’s stage of media development.
IS DOORSTROOM HET ANTWOORD? VAN BASISVAARDIGHEDEN NAAR MEDIAWIJSHEID: DREMPELS EN BEPROEVINGEN BINNEN EEN E-INCLUSIE INITIATIEF - KELIA KANIKI MASENGO AND ILSE MARIËN Het digitaal insluiten van kwetsbare individuen en groepen is een van de voornaamste hoofddoelen van e-inclusie initiatieven. Gebruik makend van laagdrempelige en op maat gemaakte methodieken trachten ze het beoogde doelpubliek te ontsluiten en te ondersteunen in het gebruik van digitale technologieën. Naast het beschikbaar stellen van computer- en internetinfrastructuur, organiseren ze eveneens leermomenten waarin kennismaken met de technologie, het verwerven van knoppenkennis en het aanleren van basisvaardigheden centraal staan. Deze leermomenten vallen onder de noemer informeel leren. De voortdurende technologische evoluties en steeds verdergaande digitalisering van de samenleving vereisen echter een kennis- en vaardigheidsniveau dat het puur operationele overstijgt. Doorstroom is een thematiek die binnen de sector beschouwd wordt als een mogelijk antwoord hierop en betreft het deelnemen aan ICT-gerelateerde opleidingen binnen het formeel onderwijsaanbod voor volwassenen. Zo ook binnen digidak, een e-inclusie initiatief dat actief is in een twintigtal gemeentes verspreid in de provincies Antwerpen en Limburg. Deze paper is opgebouwd uit twee luiken. In eerste instantie wordt de problematiek van doorstroom binnen digidak onderzocht met specifieke aandacht voor drempels tot participatie aan het regulier ICT-gerelateerde onderwijsaanbod, gekoppeld aan de perceptie van bezoekers en cursisten hierop. In tweede instantie zal de mogelijkheid tot het introduceren van mediawijsheid binnen de digidakcontext geëxploreerd worden. In het theoretisch deel van het eerste luik wordt de problematiek van doorstroom ontleed en worden achterliggende veronderstellingen bloot gelegd. De impliciete assumptie dat participatie aan volwassenenonderwijs een gunstig effect heeft op het vaardigheidsniveau van de deelnemende is aannemelijk doch niet gestaafd. Deelname aan volwassenenonderwijs wordt bemoeilijkt door een samenspel van dispositionele, situationele, institutionele en informatieve leerdrempels. Zeker in het geval van maatschappelijk kwetsbare individuen, die door negatieve schoolervaringen en een negatieve perceptie van het eigen leervermogen minder geneigd zijn tot deelname dan andere groepen in de bevolking. In het empirische luik van dit deel wordt er aan de hand van diepteinterviews gepeild naar de eigenlijke noden en behoeften van digidak-bezoekers en –cursisten inzake 55
56
ETMAAL 2013
BOOK OF ABSTRACTS
doorstroom. Het tweede luik van deze paper exploreert de introductie van mediawijsheid binnen de digidaksetting. Het begrip mediawijsheid wordt uiteengezet aan de hand van een bespreking van een aantal essentiële vaardigheden en geletterdheden waarop men beroep doet wanneer men gebruik maakt van digitale technologieën. Er wordt onderzocht welke vaardigheden onontbeerlijk zijn met betrekking tot mediawijsheid binnen de digidak-context; er worden drie mediawijze topics geïdentificeerd met name informatie zoeken, kritische omgang en privacy-issues. De levensvatbaarheid en relevantie van deze topics worden in een brainstormsessie getest, zodanig een vertaalslag gemaakt kan worden naar de praktijk. Hoofddoel is achterhalen of de drie mediawijze topics relevant zijn in het dagelijks leven van maatschappelijk kwetsbare individuen. Indien dit het geval is zullen manieren onderzocht worden op welke manier deze doelgroep bereikt en bewust gemaakt kan worden. Aan deze sessie nemen zowel mediawijsheid-experts, professionelen uit het werkveld, digidak-begeleiders als ervaringsdeskundigen in de armoede deel. De eerste resultaten van de diepte-interviews tonen aan dat er weinig belangstelling bestaat voor het doorstromen van digidak naar formele ICT-opleidingsinitiatieven. Er wordt nauwelijks een link gemaakt met deelname aan formele initiatieven en het verhogen van het eigen vaardigheidsniveau. Hoewel sommige respondenten aangeven dat ze graag willen bijleren en digitaal vaardiger willen worden, is deelname aan het regulier volwassenenonderwijs geen strategie waarvoor ze opteren. In de kantlijn dient vermeld te worden dat er binnen digidak mogelijk een Mattheus-effect kan spelen wat betreft de deelname aan de initiatiecursussen. Zo nemen aan de cursussen veelal personen deel die zich in een eerder bevoorrechte positie bevinden en die hierdoor hun positie verstevigen. De voorlopige resultaten van de brainstormsessie geven aan dat projectmatige initiatieven met ICT als zijdelingse insteek de voorkeur dragen van de stakeholders. Het idee van een groepsaankoop voor internetabonnementen werd met veel enthousiasme onthaald. Wat betreft het introduceren van mediawijsheid binnen de digidak-context werd het volgende geconstateerd: de rol van de begeleider binnen de digidak-setting is van opmerkelijk belang. Begeleiders functioneren werkelijk als katalysatoren. Inzichten uit de diepte-interviews tonen aan dat de begeleiders een aanzienlijke rol spelen in het al dan niet overgaan tot gebruik van sociale netwerksites. Inzetten op het bewustmaken van begeleiders op het gebied van mediawijsheid is een eerste stap. Het stapsgewijs mediawijs maken van begeleiders zou een trickle-down effect kunnen creëren en zo uiteindelijk ook bezoekers en cursisten kunnen beïnvloeden.
MEDIAWIJSHEID AAN DE KEUKENTAFEL. COMMUNICATIE BINNEN NEDERLANDSE GEZINNEN OVER MEDIA(-WIJSHEID) - ELIZE VAN DER MEULEN AND JORDY GOSSELT De toename van het aantal mediamogelijkheden heeft ertoe geleid dat media voor kinderen een centrale rol zijn gaan spelen in hun dagelijkse leven (Huesmann, 2007). Gezien de constante vooruitgang op het gebied van mediatoepassingen bestaat er een voortdurende zorg over mediagebruik en de effecten ervan op de jonge ontvanger (Brown & Marin, 2009). Theoretische concepten als cultivatie en modelling rechtvaardigen deze zorg. Daarom blijft het belangrijk dat kinderen op een verantwoorde manier (leren) omgaan met media. Mediawijsheid speelt hierin een belangrijke rol.
56
57
ETMAAL 2013
BOOK OF ABSTRACTS
Volgens Sixma (2007) staat mediawijsheid voor de competenties die nodig zijn op de gebieden kennis, vaardigheden en mentaliteit. Het kunnen interpreteren van mediaboodschappen, het plaatsen van communicatie binnen een historisch kader en het bewustzijn van de rol die media in het persoonlijke en maatschappelijk leven spelen zijn competenties die vallen onder ‘kennis’. Het kunnen doorzien van mediaboodschappen en het vermogen om feit en fictie te onderscheiden is daarbij van groot belang. Onder ‘vaardigheden’ wordt de omgang met techniek en informatie bedoeld. Bij ‘mentaliteit’ gaat het over het dragen van verantwoordelijkheid met betrekking tot het produceren en verspreiden van media (in het geval dat men zelf media ontwikkelt) en het hebben van een kritische houding ten aanzien van het eigen mediagebruik (Sixma, 2007). Kort gezegd is mediawijsheid een koepelcompetentie die jongeren, maar ook ouders, leert omgaan met en wapent tegen media (Bruinenberg, 2012). Omdat kinderen de meeste media binnenshuis gebruiken, is het belangrijk om rekening te houden met de rol van ouders in het begeleiden van het mediagebruik van hun kind en de rol die ouders hebben in het ontwikkelen van begrip voor media (Mendoza, 2009). Ouders kunnen kinderen helpen bij het ontwikkelen van een visie over media, bij het op waarde schatten van media en bovendien bij de plezierbeleving aan media. Binnen de literatuur worden meerdere begeleidingsstrategieën onderscheiden die ouders kunnen inzetten welke door Mendoza (2009) zijn samengevat als 'restrictive mediation' (het maken van regels met betrekking tot het mediagebruik thuis), 'active mediation' (het aangaan van een gesprek met het kind over mediaboodschappen), en 'co-viewing' (het meekijken of meedoen met het kind, zonder discussie over de media-inhoud zelf). Het doel van deze studie was drieledig. Ten eerste is onderzocht of, welke en in welke combinatie ouders gebruikmaken van de genoemde begeleidingsstrategieën. Ten tweede is onderzocht welke motieven ouders hebben voor de gehanteerde strategie(ën) en welke aspecten van mediawijsheid zij daarbij van belang achten. Ten derde zijn de percepties van kinderen met betrekking tot de communicatie binnen het eigen gezin in kaart gebracht. Om de onderzoeksvragen te beantwoorden zijn semigestructureerde focusgroup interviews gehouden bij tien gezinnen thuis, aan de keukentafel. Uiteindelijk participeerden 7 vaders, 9 moeders, 8 zonen en 9 dochters. De leeftijd van de ouders varieerde van 32-54 jaar (N=16); de leeftijd van de kinderen varieerde van 4-12 jaar (N=7) en 13-18 jaar (N=10). Aan de hand van de begeleidingsstrategieën zijn ouders en hun kinderen gevraagd naar hun omgang met de media. Na afloop van de interviews is alle data gecodeerd (Cohen’s Kappa = .73) en vervolgens geanalyseerd in Atlas.ti. Uit de resultaten blijkt dat alle drie begeleidingsstrategieën worden toegepast in de ondervraagde gezinnen en dat het multi-dimensionale aspect van ouderlijke begeleiding werd bekrachtigd door de data. Restrictive mediation wordt vooral toegepast binnen gezinnen met jonge kinderen en deze ouders verwachten meer negatieve dan positieve effecten van media. Active mediation wordt door ouders voornamelijk ingezet om hun kinderen kritisch na te laten denken over hun eigen gedrag met betrekking tot mediagebruik, maar dit wil niet zeggen dat deze ouders per definitie negatief zijn over de media-inhoud zelf. In elk gezin bestaan er regels betreffende het mediagebruik in huis, wordt er bij gelegenheid met elkaar gepraat over media-inhoud en wordt er gezamenlijk, door ouders en kinderen, gebruik 57
58
ETMAAL 2013
BOOK OF ABSTRACTS
gemaakt van media. De invulling die ouders aan deze strategieën geven verschillen echter van elkaar. Enerzijds worden de begeleidingsstrategieën door ouders toegepast om hun kinderen te beschermen tegen media-effecten waarvan zij denken dat deze schadelijk kunnen zijn. Anderzijds vinden ouders het belangrijk hun kinderen te wijzen op hun gedrag met betrekking tot mediagebruik en de rol die dit speelt in hun dagelijkse leven. De bevraagde kinderen hebben een neutrale of positieve houding ten aanzien van de ouderlijke communicatie met betrekking tot mediawijsheid, ook wanneer hen iets wordt verboden. Wel dienen ouders in deze gevallen een uitleg te geven over het waarom van het verbod of de beperking.
Session #9 Friday 8 February 09:00 – 10:00 hrs. Chair: Hilde van den Bulck Location: Lund (M1-16) English session
MUSEUMS 2.0: A STUDY INTO EXPERTISE AND CULTURE WITHIN THE MUSEUM BLOGOSPHERE JESSICA VERBOOM AND PAYAL ARORA For a long time, museums had a defining position as authorities in their respective fields. Now, under the current pressures of reduced state funding and increasing competition, the ‘participatory culture’ has become a celebrated term within the museum field. This is a paradoxical trend as it collides with the traditional high barriers of entry of these cultural institutions with their large capital of expertise. Studies on popular culture suggest that traditional experts can be seriously challenged with the entrance of the public as amateur through social media. While these studies have a more vast understanding of the impact of the participatory culture on experts and expertise in the online space, there is a dearth of literature on the high culture spectrum of the Web. We aim to answer the following research question here: who are the experts and what is the nature of their expertise in the museum blogosphere?
KNOWLEDGE AND INNOVATION IN MEDIA CLUSTERS: THE CASE OF HILVERSUM, NETHERLANDS ERIK HITTERS The main question that drives this research project is how knowledge and innovation develop in colocated media industries. The research thus focusses on 1) the development of and interrelationships between companies, markets, networks and the places where they are located (clusters) and 2) the effects of their co-location on both the companies themselves, their competitiveness, innovative capacities, collaborations, but also on the environment in which they are located. We hypothesize that physical proximity at the micro level facilitates but not necessarily generates face-to-face contacts, social networks, firm interaction and tradable and untradable transactions. These in turn might stimulate competitiveness and innovation. Our research aims to unveil which cultural, institutional and functional variables add value to media clusters.
58
59
ETMAAL 2013
BOOK OF ABSTRACTS
The focus of this paper is on the cluster of media industries in the city of Hilversum, the Netherlands. Hilversum is a medium- sized town of some 84,000 inhabitants, but it is home to most of the television and broadcast- related companies and institutions in the Netherlands. About a third of all jobs in this city are related to the media. The focal point of the broadcasting cluster is the Media Park, a business park almost exclusively for media businesses, that houses some 200 companies. Although broadcasting is the main focus of the Hilversum media cluster, it is also a city that has a strong presence in adjacent fields of multimedia, ICT and creative industries. The cluster is, however, losing ground to its main competitor in this industry, Amsterdam. Until the beginning of the 21st century, the broadcasting industry has grown steadily in Hilversum, but in the last decade this growth has slowed down, and employment has been falling since 2008 (Immovator, 2010). The project uses a mixed approach of both extensive quantitative, and intensive qualitative information. The extensive quantitative data will draw on several data sources such as statistical and economic data offered by LISA (the Laboratory for Interdisciplinary Statistical Analysis). I will also use secondary analyses of a number of reports on Dutch media, such as the Cross Media Monitors (Immovator, 2006, 2008 and 2010). The paper adds interview-based qualitative data, from 21 interviews with media companies located in Hilversum, that may unveil hidden practices of innovation and knowledge transfer.
WHAT IS COOLNESS? DECONSTRUCTING AND CONTEXTUALIZING A BUZZWORD - JANNA MICHAEL Coolness is an almost omnipresent construct in popular culture: Advertisements strive to convey the coolness of products, jury-members label contestants of singing and dancing competitions as ‘cool’ and violent youngsters are often obliged to fulfill ‘coolness’ trainings. From this, we can already sense that there are several different ideas on what coolness means. While coolness is used in various ways in everyday-life, pin-pointing it and defining it in clear terms is considered rather complicated for scholars. Thinking back to the early days of jazz, musicians were the ‘inventors of coolness’. Their cool-pose has been described as the outcasts’ chance to be admired. In line with this, coolness has been described as a game that everybody can play independently of someone’s economic or cultural resources. ‘Winning the game’ in this sense means gaining social status or recognition in a group. Nowadays the buzzword coolness is used inflationary for expressing admiration of basically anything and anybody. This study disentangles different understandings of coolness and further aims at creating a closer understanding of its current relevance as a potential form of capital. It will do so by discussing the different notions of coolness in the literature and contrasting these with an analysis of semi-structured interviews with young members of the creative class. On basis of the interviews, I will show how coolness can be meaningfully linked to symbolic and subcultural capital. Consequently this analysis will deconstruct coolness and evaluates its importance for current discussions on cultural hierarchies.
59
60
ETMAAL 2013
BOOK OF ABSTRACTS
Session #10 Friday 8 February 10:15 – 11:15 hrs. Chair: Stijn Reijnders Location: Lund (M1-16) English/Dutch session
NEW MEDIA PROSUMPTION IN THE GLOBAL SOUTH - PAYAL ARORA AND NIMMI RANGASWAMY Photoshopping of newlyweds, downloading the latest movies, teens flirting on social network sites and virtual gaming may seem like typical behavior in the West; yet in the context of a village in Mali or a slum in Mumbai, it is seen as unusual and perhaps an anomaly in their new media practice. In recent years, some studies (Ganesh, 2010; Arora, 2010, 2012; Rangaswamy, 2009; Kavoori, Chadha & Arceneaux, 2006) have documented these leisure-oriented behaviors in the global south and argued for the need to emphasize and reposition these user practices within larger and contemporary discourses on new media consumption. Yet, for the most part, studies in the field of Information and Communication Technologies for Development (ICT4D) have duly relegated such enactments as anecdotal. This is partly due to the fact that much of this research is driven by development agendas with a strong historical bias towards the socio-economic focus (Burrell & Anderson, 2009). Data that is not directly addressing project-based outcomes is sidelined. However, as emerging economies globalize and urbanize exponentially, and their users become more critical consumers and creative contributors of digital content or ‘prosumers’ (Bruns, 2008) and arguably free laborers (Scholz, 2012) instead of classic development beneficiaries, a paradigm shift is needed in approaching this new media audience with a more open-ended, explorative and pluralistic perspective. Thereby, this article serves as a call to rethink new media practices in the global south by looking at the implications and impacts of ICTs as leisure (entertainment/pleasure/ play) artifacts in the context of developing economies and emerging markets. We believe this line of inquiry is timely and enables a strategic bridging of the new media studies and development communication domain. Despite studies yielding insightful commentaries on ICTs in this arena, we believe resource constrained environments generating rich usages that are not overtly utilitarian have remained hitherto unexplored. A critical movement is needed among scholars focusing on emerging economies to reconceptualize the mobilization and serviceability of ICTs to extend beyond a conservative understanding of developmental value. This will help in focusing on the heterogeneous and life enhancing aspects of technological use encompassing both experiential and purposive elements of ICT adoption: their interplay with systematic/systemic ecological constraints to provide an analytical and descriptive study of the technology spectrum and use in these contexts. To illustrate our argument, we conduct an extensive comparative and sociological analysis of a range of ethnographic studies across the emerging economies to outline the range of ways in which leisure is being enacted within these novel contexts. We embed these practices under four dimensions of critique and we recommend four strategic ways in which we can open up these conversations in the development communication domain: a) broadening the development lens for technology adoption b) ICTs are social artifacts before they become tools for development, c) Privileging individual 60
61
ETMAAL 2013
BOOK OF ABSTRACTS
aspirations/mannerisms over communal development and, d) Sidelining morality and virtuosity of the poor for pluralistic possibilities. To summarize, we offer some critical points of contention that need addressing and new avenues for research if we are to rethink, reframe and refresh our thinking on Web 2.0 enactments in the global south.
WATCHING TV FICTION IN THE AGE OF DIGITIZATION - NELE SIMONS Since the birth of broadcasting television, Television Studies scholars have been studying the meaning of television in daily life and television viewing experiences. However, over the last decade, the nature of our television experience has changed significantly. Only a few years ago, televisual narrative was restricted to a certain kind of content from a limited number of sources via a television set; now audiences face a wide array of choices in how, when and where they engage with it (Evans, 2011). The questions then becomes, how are these televisual narratives being consumed by viewers? What choices are audiences making and why? The set broadcasting schedule of traditional television determines when we watch (time), what we watch (content) and where we watch (location) (Van den Broeck, Pierson, & Lievens, 2007). With the digitization of television the interrelation between these three dimensions disappears. Due to the fixed and repetitive programme schedule, traditional broadcasting television serves as a sort of timetable. People’s everyday activities are shifted, elongated or cut short to accommodate the programs they want to watch (Gauntlett & Hill, 1999). The program schedule is controlled by the broadcaster. This time structuring element is challenged by time-shifting technologies, such as digital television (DTV, often with recording function), downloading, DVD. Although these are three very different technologies, they all enable a more random-access mode of viewing fiction series. Second, there is the content dimension. Traditional linear television can also be described as a push environment, meaning that the audience remains passive as the television programmer decides what content is pushed to the viewer. Through digitization, viewers cannot only manage when and where to watch, but also what they watch. The viewing environment evolves from a push environment into a pull environment, meaning that the viewer can pull the desired content. Third, the fact that television is heavily integrated in people’s daily lives and routines is also translated in the central place of television in the living room. Although today many households have more than one television set, the television set in the living room is usually the best equipped one (Bauwens, 2002). The process of digitization has detached television content from the television screen. Not only is television content available on the different television sets in the house, it is also available on computer screens, mobile phones and other portable devices. Based on the technological possibilities of digital television, one would expect a radical change in audiences’ television viewing practices: they are now able to choose when (time), what (content), and where (location) they want to watch. Studies making these predictions about changing viewing practices often tend to focus on the textual and industrial changes, making assumptions about technological possibilities. However, little research has focused on the impact of these changes on the actual viewing practices of audiences and the reasons behind these practices. We believe it is necessary to examine how audiences are embracing these changes. Through qualitative audience research, using television diaries and in-depth interviews, we studied people’s TV fiction viewing practices. Our sample consisted of 61 engaged TV fiction viewers, meaning that these viewers do 61
62
ETMAAL 2013
BOOK OF ABSTRACTS
more than just watching episodes through live television broadcasting. While doing the analysis it became clear that certain tendencies in viewing practices could be discovered. These trends come together in viewing practices profiles (see below). Not every respondent in this study can be categorized in one of these profiles, and respondents are not permanently assigned to one type of viewing. Rather, these profiles are prototypes of viewers on a continuum of viewing practices. “The live viewer”: watches live on the TV screen, often with others, one episode at the time, usually Flemish fiction. This viewer finds it convenient that the broadcaster is in control. “The DTV viewer”: watches delayed on the TV screen, often individually, one episode at the time, both Flemish and international fiction. This viewer likes to be in control of the TV schedule, but is happy to wait for the Flemish release of international fiction. “The downloader”: downloads all TV fiction and watches the episodes on the TV or laptop screen, usually individually. This viewer watches international fiction immediately after the original release. “The DVD viewer”: waits until the whole series is released on DVD and watches multiple episodes in one sitting on a TV or laptop screen, alone or with others. This viewer likes to be in control over the rhythm of watching.
MEDIAGEBRUIK, BODY DISSATISFACTION EN APPEARANCE ANXIETY: EEN STUDIE BIJ NEGEN- TOT TWAALFJARIGE MEISJES - SASKIA MERRE AND STEVEN EGGERMONT De afgelopen decennia is de relatie tussen mediagebruik en het lichaamsbeeld van meisjes en vrouwen steeds vaker de focus geweest van onderzoek. Verscheidene studies toonden aan dat televisie, magazines en het internet een invloed uitoefenen op de ontwikkeling van het lichaamsbeeld bij meisjes en vrouwen en dit door de frequente aanwezigheid van het slanke schoonheidsideaal (e.g. Jones, Vigfusdottir & Lee, 2004; Tiggemann & Slater, 2003; Tiggemann & Miller, 2010). Ondanks de toegenomen interesse in dit domein, blijkt echter weinig aandacht geschonken te worden aan meisjes in de pre-adolescentie, d.i. tussen negen en twaalf jaar oud. Bovendien ligt de focus in deze schaarse studies vaak enkel op body dissatisfaction, waardoor de onzekerheid en angst die meisjes ervaren over hun uiterlijk – appearance anxiety – uit het oog wordt verloren. Deze lacunes zijn belangrijk, omdat onderzoek aantoont dat meisjes in deze leeftijdscategorie reeds intensief begaan zijn met hun lichaam en uiterlijk, en ook al goed op de hoogte zijn van verscheidene dieetpraktijken (Clark & Tiggemann, 2008; Dohnt & Tiggemann, 2006; Phares, Steinberg & Thompson, 2004). Bovendien wordt een negatief lichaamsbeeld gekoppeld aan verscheidene psychologische en fysieke gezondheidsproblemen, waaronder een lagere self-esteem, depressiviteit, dieetgedrag en eetaanvallen (Aubrey, 2006; Holsen, Kraft & Roysamb, 2001; Stice, Marti & Durant, 2011). Dit zijn problemen die steeds vaker bij meisjes in de pre-adolescentie worden vastgesteld, wat de hypothese voedt dat processen die frequent geobserveerd worden in de adolescentie reeds starten in de periode die daaraan voorafgaat. Het huidige onderzoek speelt dan ook in op bovenstaande beperkingen door – vanuit het Tripartite Influence Model (van den Berg e.a., 2002) – na te gaan wat de mogelijke impact is van mediagebruik op body dissatisfaction én appearance anxiety en dit bij een jongere leeftijdsgroep, namelijk Vlaamse negen- tot twaalfjarige meisjes of tweens. Aan 524 meisjes uit het Vlaamse onderwijs werden vragenlijsten voorgelegd die onder andere 62
63
ETMAAL 2013
BOOK OF ABSTRACTS
peilden naar hun sociaal-economische achtergrond, algemeen televisie- en tijdschriftgebruik en lichaamsbeeld, bestaande uit body dissatisfaction en appearance anxiety. Verder werden bijkomende beïnvloedingsprocessen – internalisering en sociale vergelijking - in kaart gebracht, aangezien voorgaand onderzoek aantoonde dat het lichaamsbeeld bij tienermeisjes en volwassen vrouwen voornamelijk indirect tot stand kwam, en dit via deze processen (Halliwell & Dittmar, 2004, Rodgers e.a., 2011; van den Berg e.a., 2002). De resultaten geven aan dat, vergeleken met meisjes in de adolescentie, de meisjes in deze studie over het algemeen tevreden waren over hun lichaam (body dissatisfaction) en relatief weinig bezorgd waren over hoe anderen hun lichaam en uiterlijk zouden kunnen percipiëren (appearance anxiety). Ondanks deze positieve bevindingen bleken de body dissatisfaction en appearance anxiety van de meisjes in deze studie toch significant samen te hangen met blootstelling aan boodschappen en beelden op televisie en in tijdschriften, en dit op eenzelfde manier als bij oudere meisjes en volwassen vrouwen. Zo bleken ook meisjes in de pre-adolescentie een significant negatiever lichaamsbeeld te ontwikkelen wanneer ze werden blootgesteld aan televisiegenres (zoals komedie, soap opera en muziektelevisie) en tijdschriftonderwerpen (zoals gezondheid, schoonheid en mode) die de nadruk leggen op het slanke schoonheidsideaal. Daarenboven bleken ook totale televisieblootstelling en totaal tijdschriftgebruik voorspellers te zijn voor body dissatisfaction en appearance anxiety, hetgeen contrasteert met voorgaande studies (Calado e.a., 2011; Tiggemann & Pickering, 1996; Tiggemann, 2003). Al de vastgestelde relaties – zowel voor specifieke als totale blootstelling – bleken bovendien, net zoals bij oudere respondenten, te verlopen via de mediatoren internalisering en sociale vergelijking. Dit wijst er op dat de tweens voornamelijk op een indirecte manier een negatiever lichaamsbeeld ontwikkelen, door zich enerzijds te vergelijken met schoonheidsidealen in de media en deze idealen anderzijds voorop te stellen als persoonlijk doel of standaard. De discussie gaat dieper in op de noodzaak om onderzoek te blijven voeren binnen deze leeftijdscategorie en focust ook op nieuwe invalshoeken voor toekomstig onderzoek.
“FREE YOUR …” “MOST OPEN” ANDROID: CRITICAL ANALYSIS OF DISCOURSES ON ANDROID LELA MOSEMGHVDLISHVILI AND JEROEN JANSZ The debate about how software is developed and distributed (whether open source or proprietary) and the merits of user rights (in terms of their freedom and control over devices) has thus far been centered on PCs and the Internet. However after the introduction of smartphones, pocket-sized computers which are programmable phones, these issues became pressing topics with political overtones also Smartphone domain. A primary point of contestation in this debate is the notion and practice of intellectual property in relation to code/software and the rights of a user. The aim of the paper is to contribute to the field by exploring how antagonism between open source (as commodified by private sector) and free¬¬/libre software (as manifested in the aims of a movement that challenges dominant relations of software production) is played out in this new domain on a discursive level. As a research focus, the most widely spread operating system for smartphones Google’s Android was selected. For clarity, it is important to note, that the initial dichotomy of proprietary vs. free/open-source software (for example in Microsoft Windows vs. Linux) has been challenged by the companies who have appropriated open-source development into new business models, for example by releasing the 63
64
ETMAAL 2013
BOOK OF ABSTRACTS
source code, but under limited conditions and restricted rights. Commodification of open source initiatives further increased the ideological gap between two movements, roughly grouped around the Open Source Initiative (see Raymond, 2001) and Free/Libre Software Movement (Stallman, 2003). Both movements aimed for openness of platforms and the release of source code, but drawing on different values and philosophies. Namely, the Open Source Initiative based on values of neo-liberalism and technical efficiency, and Free/Libre Software utilized a discourse of democratic rights and freedom of users (Berry, 2008). Currently proprietary vs. open source software does not represent dichotomy, since open source has been integrated in the dominant modes of software development, leaving free/libre software as an alternative movement. Initially, Android was indeed based on open source and free/libre Linux Kernel, and developed by a small startup group. However when Google acquired Android, it implemented a small but a crucial change in software licensing. Linux was protected by the most widely used free software license the GPL (the General Public License), whereas Google’s Android (more precisely all parts that are not Linux Kernel) are now being released with the Apache license, that is open source but not free/libre license (Authors, 2012). As a response to this, the Free Software Foundation Europe (FSFE) launched a campaign “Free Your Android” to raise users’ awareness regarding the limitations of Android (e.g. practice of gathering and monetizing user data without a person’s consent, the right to contribute directly to operating system and not merely write an App for it). The FSFE claims that although Android is “less bad” than other Smartphone platforms (namely proprietary Apple’s iOS or Microsoft’s Windows Phone) it still does not respect user’s rights; and is not free/libre software. Free in this context means free as in ‘freedom of speech’, and not as in ‘free beer’. The study critically analyzes discourses on Android by Google and the Free Software Foundation Europe (FSFE) and aims to reveal the differences in conceptualizing notions such as open-source and free software. The leading research questions are (a) how do discourses on Android differ between these groups, (b) what meanings do words such as ‘free’ and ‘open’ carry in their communication, (c) how do the discursive differences effect identities of these groups. Feenberg’s (2010) Critical Theory of Technology provides the theoretical backbone of this study because it rejects technological determinism and emphasizes user agency, that is, the user’s role in technological decision-making. Empirical data for the study comprises Google’s official communication on Android (as manifested in press releases issued by Google’s corporate press center in the period of 2007-2012) and FSFE materials (press releases, campaign texts, newsletters). A combination of Laclau and Mouffe’s discourse theory (2001) and Critical Discourse Analysis (Fairclough, 2010) is used to analyze textual material. Preliminary findings reveal a vivid conflict between Google and the FSFE. Whereas Google emphasizes the open nature of Android, its use of the word free radically differs from the meaning that FSFE ascribes to it. In their communications on Android, the concept open source is been neutralized within the hegemonic order of discourse on intellectual property, leaving Free/Libre Software as an oppositional discourse. 64
65
ETMAAL 2013
BOOK OF ABSTRACTS
Session #11 Friday 8 February 11:45 – 12:45 hrs. Chair: Stijn Reijnders Location: Lund (M1-16) English session
EXPLAINING GLOBALIZATION IN POP MUSIC CHARTS. A MULTILEVEL ANALYSIS OF NINE COUNTRIES BETWEEN 1960 AND 2010 - MARC VERBOORD AND AMANDA BRANDELLERO Over the past decades we have witnessed a clear increase in the exchange of cultural products across the globe. This has manifested itself not only in terms of trade, but also with regard to cultural gatekeepers discussing foreign arts and culture (Janssen, Kuipers & Verboord, 2008) and artists recording abroad (Brandellero and Pfeffer, 2011). Trends regarding consumer preferences have received less attention and have tended to focus on a limited number of countries and cultural products, for instance television programmes and films. The globalization of pop music and its translation in charts performance remains relatively unexplored to date. The few studies which have examined trends in the demand for foreign music have emphasized parallels with developments in the international political arena (Achterberg et al., 2011), rather than trends in the fields of media and culture. Yet, the rise of commercial broadcasters, music television (most notably MTV), street cultures (e.g. hip hop), and the blurring boundary between celebrity and artistic talent (e.g. Idol-type TV formats), in addition to the size of the national music industry and the general level of receptivity of a country to foreign cultural products, have affected the ways audiences, in the course of time, discover and appreciate music. We argue that these contextual factors help to explain cultural globalization as they can be interpreted as instrumental to processes of ‘cultural imperialism’, ‘cultural flows’ or ‘glocalization’. In pop music, some countries in particular occupy more central positions in global cultural production (US and UK) and are, consequently, more able to dominate the field. In the early days of MTV, American and British acts benefitted strongly from this position; in the late 1990s when regional versions of music television emerged, this advantage decreased. Despite the expected dominance of English-language pop music, both imported and locally-produced, other flows of music are also expected to occur. Research into television and film has shown how cultural proximity, for instance in language and other cultural traits, influences the supply and demand of foreign products. We anticipate similar dynamics at work in the diffusion of genres and artists rooted to local music scenes. Moreover, culture is often also ‘made’ more similar: cultural entrepreneurs adapt existing genres – often from countries with central positions – to more local versions. Empirically, this paper uses weekly music chart data from 9 countries in the period 1960-2010 to study the extent to which consumers of pop music have become globally oriented in the past half century. Our explanatory analyses comprise country level data (degree of cultural centrality of music industry, cultural proximity, media systems, foreign language proficiency, GDP, population size) and artist level data (genre, language, star power) to explain processes of cultural globalization. We 65
66
ETMAAL 2013
BOOK OF ABSTRACTS
advance on previous work by (a) analyzing not only Western European countries and the US, but also North and South European countries and Australia, (b) increasing reliability of our study by including also smaller hits, and (c) broadening the time frame. We conduct a multilevel analysis to assess the impact of different factors. The results show just a little convergence toward globalization over time. While countries like Italy and France have more and more opened up to foreign products, the United States remain highly focused on domestic content. There are considerable fluctuations, however, depending on fads in musical genres (British beat wave in 1965-1970, German, Dutch and Italian eurohouse in the 1990s), and media developments (MTV stimulating internationalization in the 1980s, American Idol-type of formats stimulating domestic music in the 2000s). The first multivariate analyses show that cultural centrality – the degree to which a country’s music is successful in abroad – has a positive effect on global success for countries of origin, but a negative one for countries of destination. That is, getting in the charts in “central” countries like the United States and the United Kingdom is difficult, but artists coming from these countries have higher chances to get global success.
THE DISCOURSE OF REGIONAL POPULAR MUSIC (RE)CONSIDERED - SIMONE DRIESSEN This paper explores the relationship between popular music and its audiences, particularly how regional popular music plays a role in daily life and what significance audiences assign to it. The study positions itself in the fields of cultural identity, audience and popular music studies, and is part of a broader project that explores music histories in the Netherlands from local, regional, national and transnational perspectives. This paper highlights the regional perspective by offering an understanding of how regional popular music fulfills a role in the everyday lives of audiences, using the Dutch province of Limburg as a setting for this case study. Limburg, the southernmost province of the Netherlands offers an interesting case study, for not only its (peculiar) dialect, but also because of the strong Catholic roots and accompanying traditions. These elements are ‘atypical’ for the Netherlands; for the largest part of the country builds on Protestant roots and the number of (regional) dialect-speaking people is highest in Limburg (still 70% is able to speak the dialect). Besides these elements, Limburg is also the region where population ageing strikes hardest and where many young adults leave the province because of the high unemployment rate. However, it can be claimed that due to the dialect and typical Catholic traditions like carnival, young adults maintain a strong relationship with the province. In order to see what significant value `Limburgians’ attach to music and how regional, popular music is given meaning in their everyday life, interviews were conducted with fourteen Limburgians who live(d) in the province. Half of the interviewees are dialect-speakers; all of them understood the dialect. The interviewees came from different age groups and educational backgrounds. Drawing on semi-structured interviews, this qualitative study offered insights in how this group of Limburgians negotiated their connectedness to the locality of Limburg; defined regional, popular music; what they considered typical for the region; how this style of music fitted in with their everyday preferences, and how it linked to their personal taste and musical practices. 66
67
ETMAAL 2013
BOOK OF ABSTRACTS
Qualitative analysis of the interview data finds that regional popular music for this particular group can be considered a ‘binding’ element that fulfills two main roles in the everyday life of Limburgians. On the one hand, regional popular music serves as a means of temporary identification with their regional identity, which the interviewees strongly linked to the setting of (the tradition of) carnival. Moreover, the interviewees consider both period and tradition of carnival as a key platform on which this particular mode of regional-popular music and their everyday lives can converge. In this sense it can be argued that it acts as an (temporary) affirmation of their regional identity and regional consciousness. On the other hand, regional popular music serves as a ‘resource of recollection’ for the interviewees; the Limburgians use regional popular music in order to evoke (or affirm) memories of the province, for the songs often refer to the landscape or lifestyle of the province. This mechanism especially applies to the group of interviewees who (for a period) left the province. However, in both cases, the dialect - in which the songs are often sung – is not necessarily a precondition for the interviewees, but it is of strong added value to the other aforementioned elements. Though, according to the interviewees it is very important that the songs should draw on shared characteristics of the Limburgian community before they consider them ‘regional’. In numerous studies is argued that music has a bonding effect on people, because it can be considered a form of shared, cultural heritage which serves as a basis for creating a collective identity. This could lead to a feeling of attachment or inclusion, which contributes to the idea of the ‘imagined community’, a community drawing strongly on a feeling of connectedness. This study concludes that regional popular music is indeed a binding element, as long as it draws on or relates to shared common characteristics. In that sense it affirms the regional identity/ regional consciousness of the Limburgians, and does it serve as a means of reminiscence of the province. Therefore music can - symbolically - fulfill a great role in an age where personal connections are believed to be decreasing at a fast pace.
GRUNTING ALONE: ONLINE GENDER INEQUALITY IN EXTREME METAL MUSIC - JULIAN SCHAAP AND PAUWKE BERKERS Popular music plays an essential role in the everyday lives of adolescents. However, both the consumption and production of popular music are highly stratified along gender lines. In this paper, we examine the production of metal music to demonstrate how participants use online social media to contest or maintain existing offline gender boundaries. Metal music is a form of cultural expression which is both numerically and symbolically male-dominated. Not only is metal music often constructed as form of male rebellion (Frith & McRobbie, 1990), previous studies have shown that learning to play in a band is largely a peer-based – rather than individual – experience, shaped by existing sex-segregated friendship networks (Bielby, 2003; Clawson, 1999). As a result, few women have successfully pursued a career in this ‘masculine’ genre (Kahn-Harris, 2007; Krenske & McKay, 2000). Their token status is likely to make women highly visible within the metal scene, to be evaluated on the basis of their sex category than their actual performance, and to receive harsher criticism compared to their male counterparts (Moss Kanter, 1977; Ridgeway, 2007; Roth, 2004). Online however, interactive social media have created new modes of social conduct that might contest (or uphold) existing gender inequality. Here we examine the performance of ‘vocal covers’ on 67
68
ETMAAL 2013
BOOK OF ABSTRACTS
YouTube which allows women (and men) to produce metal music without needing a ‘real’ band. On the one hand, social media may facilitate women to actively engage in the metal scene. On the other hand, female participation might be limited to the virtual scene without allowing them to kick-start an offline music career as well. Few studies have addressed the (gendered) relationship between online and offline music scenes (Bennett & Peterson, 2004). This paper therefore addresses how females (and male) vocalists navigate the virtual metal scene, and how this relates previous research on offline gender dynamics. To answer our research question, we have conducted a quantitative and qualitative comparative content analysis of YouTube user-comments (utterances) of males and females performing vocal covers (n=972). First, all utterances were analysed using three main categorizing variables (type of utterance, topic of utterance, evaluation in utterance). Second, two variables that measure online gender dynamics in the comments (gender-biased evaluation, breaking of gender roles) were applied. An example of a gender-biased evaluation is that a commentator thinks the female performer is good, ‘for a girl’, implying an evaluation based on group category rather than individual skills. An example of the breaking of gender roles is when a commentator did not expect a woman to perform (because of the aforementioned role encapsulations) and subsequently reveals surprise, anger, happiness etc. as a reaction to it. Third, a final variable measured whether an utterance included either a romantic love gaze or an erotic (or sexual) gaze. Our analysis revealed that in line with previous research women are also highly underrepresented in the online metal scene, at least as producers of vocal covers. Surprisingly however, we did not find many differences in the actual content of the comments male and female vocalists received, especially considering positive and negative judgements about the performance. Utterances containing a bias along gender lines or the breaking of gender roles revealed that females do receive more comments that address their status as tokens within the metal scene. This verified their higher visibility and their minority status within the scene. As tokens, females were expected to receive more evaluations based on their appearance and non-ability traits than males (Johnson-Grau, 2002), but our analysis revealed that males receive as much comments addressing their appearance as females.Women are however significantly more often the subject of either a romantic love gaze or an erotic gaze than male performers. However, comments that verify the breaking of gendered role encapsulations or that are biased along gender lines are in many cases positive rather than negative, implying that the online scene might be utilized to circumvent offline gender dynamics and consequently might serve as a springboard to an artistic career in the offline metal scene.
REMEMBERING THE LONG SIXTIES: A EUROPEAN POPULAR MUSIC HERITAGE? - ARNO VAN DER HOEVEN The aim of this paper is to explore the memories of popular music from the 1960s from a European perspective. Numerous publications have shown how rock-n-roll and beat music had a pivotal role in the then emerging youth culture and countercultural activities of protest movements in various European countries. Moreover, music came to symbolize the cultural styles from the sixties and the advent of a youth oriented consumer culture. This paper examines how these developments are remembered within a European context, focusing on the different actors involved in shaping music related memories. 68
69
ETMAAL 2013
BOOK OF ABSTRACTS
Research on European identity finds that although identification with Europe is widespread, national identities remain very powerful. However, most studies approach European identity from the perspective of civic and political identities. This paper, in contrast, explores to what extent European identity is rooted in a shared cultural past. While there is a considerable amount of studies showing the importance of popular music for imagined communities on local and national levels, less is known about the relation between European identity and popular music. In order to examine the role of popular music in the construction of a European memory culture, this study draws upon the theoretical distinction between memory, history and heritage. Memory refers to the lived experiences of European citizens. History, however, assumes a more distanced and intellectual orientation towards the past. Thus, history often shuns away from nostalgia and received narratives. Heritage, finally, is about a collective appropriation of memories. Consequently, shared memories are recognized as constituting a collective identity. The findings of this article are based upon in-depth interviews with professional and amateur heritage practitioners. Furthermore, examples of popular music heritage such as documentaries, websites, archives and exhibitions are examined. The analysis of this material suggests that memories of the baby boomer generation transcend national boundaries and particular narratives of the sixties are shared in many countries. In relation to this, the paper discusses historical studies which compared experiences of the sixties across European countries. While these studies have debunked various myths of the sixties and have shown significant differences between countries, they corroborate the transnational character of popular music and youth movements from the sixties. However, as the findings of this article indicate, this does not mean that these events are necessarily remembered as European cultural heritage. Museums, archives and media commemorating the sixties, generally target audiences in particular countries. Thus, this paper highlights a tension between what has happened in Europe according to historic studies of the sixties and the ways in which these developments are remembered.
69
70
ETMAAL 2013
BOOK OF ABSTRACTS
HEALTH COMMUNICATION Session #1 Thursday 7 February 14:05 – 15:05 hrs. Chair: Anika Batenburg Location: Praag (M3-05) English session
HEALTH ANXIETY, MEDICAL CONSULT SATISFACTION AND THE ROLE OF INTERNET INFORMATION MARTIN TANIS There is a body of literature that describes how people with increased anxiety about their health make use of medical care. Health anxiety refers to the misinterpretation of all kinds of bodily symptoms that may indicate illnesses with hypochondriasis at its extreme form. Research has shown that health anxious individuals visit doctors more often, expect to get referred to a specialist more often, and in general make more use of all kinds of health care. It is also known that health anxious individuals perform all kinds of safety seeking behavior in order to protect their health or to reduce fear and uncertainty regarding their health situation. Finding information on the internet can be seen as one of these safety seeking behaviors. Research has shown that health anxious individuals differ in the way they make use of the internet compared to non-health anxious: Health anxiousness is related to longer, more frequent, more elaborate and more active (for instance in the form of posting messages on forums) seeking for health related information, and maybe even more importantly, a more negative response towards the retrieved information. The present study seeks to explain how online information seeking might effect health anxious patient’s satisfaction with their doctor’s visit. Findings are based on a survey conducted among patients prior and after their doctor’s visit (N = 185). Results show that health anxious individuals more often draw the wrong conclusions about their health situation (r = .19, p < .05). Drawing the wrong conclusion is negatively related to satisfaction with the consult (r = -.22, p < .05). More importantly, the relation between health anxiety and drawing the wrong conclusion is partly mediated by seeking online information prior to the consult. In other words, health anxiety causes increased searching for information about one’s own medical condition, which affects the correctness of one’s conclusion about a health issue. Results are discussed by looking at the cognitive behavioral model by Salkovskis and Warwick (1986) that provides us with insights in how health anxious individuals may use the internet for information. According to this model health anxious information take in information in a biased manner. The assumption is that health anxious individuals have a preconceived assumption that bodily symptoms signal that something is wrong with the body, and that a (more or less) severe illness is likely. As a consequence, people may become hyper sensible for these symptoms. In order to limit their worries, 70
71
ETMAAL 2013
BOOK OF ABSTRACTS
they will try to get certainty, for example by seeking information on internet. However, according to the model, the information that is found will be processed in a biased way. Health anxious individuals tend to have selective attention to information that confirms that they have a certain illness, and in addition, tend to ignore information that signals good health. All in all, when looking for information, health anxious individuals often tend to focus only on the catastrophic side of the situation, which in tern may increase fear for having a serious medical condition. Even though the present study does not provide us with information about what the patients have actually found or remembered on the internet, it would provide a sound explanation for the results found in this study.
HEALTH INFORMATION SEARCH OF NATIVE-DUTCH AND TURKISH-DUTCH GP PATIENTS AND ITS RELATION WITH PATIENT PARTICIPATION - SANNE SCHINKEL, JULIA VAN WEERT, JORRIT KESTER AND BARBARA SCHOUTEN Introduction: Previous research indicates that ethnic minority patients are less participative (e.g. asking questions, relative talk) than patients from the majority group (Cooper-Patrick et al., 1999; Johnson, Roter, Powe & Cooper, 2004). This passive behavior of ethnic minority patients is unfortunate since patient participation is a key factor for effective medical communication and positive health outcomes (Street, 2011). The lower participation of ethnic minority patients could be due to a difference in health information seeking behavior before the consultation, because obtaining health information before the consultation is considered an important determinant of patient participation (Cahill, 1998). Informed patients better understand treatment issues (Street, Voigt, Geyer, Manning & Swanson, 1995) and hence, are better equipped to discuss the problem and making health decisions. As there is no research yet testing the above-mentioned propositions, the aims of this study are to gain insight into: 1) differences between native-Dutch and Turkish-Dutch patients in information seeking behavior before the consultation and patient participation during the consultation and 2) the extent to which information seeking behavior before the consultation affects patient participation. Methods: Patients were recruited in six practices in three multicultural cities in the Netherlands with eleven GPs (seven men, four women). Two patient samples were used: sample one completed a preconsultation questionnaire, sample two completed the same pre-consultation questionnaire and their consultation were audiotaped. Total sample to analyse differences in information seeking behavior before the consultation consisted of 82 native-Dutch and 74 Turkish-Dutch patients (N = 156). From sample two (82 native-Dutch, 38 Turkish-Dutch patients), 34 native-Dutch and 34 TurkishDutch patients were matched on age and gender, since the groups significantly differed on these background characteristics. This sample is used to measure the participation during the consultation and the relationship between information seeking beforehand and patient participation. Results: Results show that both patient groups favor the Internet for searching health information. Turkish-Dutch patients consult books, magazines and newspapers and television and radio more often than native-Dutch patients. Interpersonal communication with family and friends on the health problem is popular in both groups. Turkish-Dutch patients are more likely to talk to companions (people with similar health problems) than native-Dutch patients. The groups report different motives for seeking information before the consultation. Turkish-Dutch 71
72
ETMAAL 2013
BOOK OF ABSTRACTS
patients report understanding the problem as most important, while native-Dutch patients mainly search information to judge whether or not to visit the GP. Both native-Dutch and Turkish-Dutch groups value the two other motives (i.e. discussing with GP and decision-making) lowest. Native-Dutch patients initiate a higher proportion of utterances on the decision and ask more questions during the consultation than Turkish-Dutch patients. Furthermore, during the consultation, native-Dutch patients refer more often to information they found beforehand and to the source of that information than Turkish-Dutch patients. Turkish-Dutch patients who report to seek for information in media are more participative than Turksih-Dutch patients who did not use media to seek for information. Discussion: Turkish-Dutch patients use a wider variety of media for seeking health information before the consultation than native-Dutch patients and seek for information due to other motives than native-Dutch patients. Despite their wider variety of media use, native-Dutch patients are more participatory during the consultation than Turkish-Dutch patients. Because Turkish-Dutch patients who sought for information in media were more participative than Turkish-Dutch patients who did not, information search does seem to be related to participation among this patient group. The finding that native-Dutch patients more explicitly refer to the information found could be explained by different communication styles between the groups. Turkish-Dutch patients are originally from a more collectivistic culture with more power distance than native-Dutch patients (Hofstede, 1984). Hence, Turkish-Dutch patients are possibly not seeking for information for their own benefit but for a more holistic understanding of the problem, and still want to leave the decision making to the doctor. Native-Dutch patients are possibly searching for their own benefit and thus will act more on the information sought. This relation needs more investigation in future studies on the impact of cultural issues on doctor-patient communication. As this study indicates that encouraging ethnic minority patients to seek for health information can benefit patient participation, it is recommended to train general practitioners to stimulate ethnic minority patients’ search for health information, thereby enhance their participation levels during medical consultations.
TAILORING ONLINE HEALTH MATERIALS TO IMPROVE MESSAGE PROCESSING AND COMPREHENSION: WHAT STRATEGY WORKS BEST FOR PEOPLE WITH LOW HEALTH LITERACY? CORINE S. MEPPELINK, EDITH G. SMIT, BIANCA M. BUURMAN AND JULIA C.M. VAN WEERT In today's society, the Internet is an important health information source. About half of the people that frequently use Internet, search for health-related information regularly (Statistics Netherlands, 2012). However, questions can be posted concerning the comprehensibility of this information, especially for people with lower health literacy. Health literacy is related to individuals’ general literacy (Sørensen et al., 2012), and can be defined as “the way in which people have the capacity to obtain, process and understand basic health information and services needed to make appropriate health decisions” Ratzan & Parker, 2000). Accurate Dutch statistics on health literacy are lacking, but approximately 40% of the Dutch participants of the Adult Literacy and Life Skills Survey did not meet level 3 (Fouarge, Houtkoop, & Van de Velde, 2011). Level 3 indicates a literacy level which is considered a “suitable minimum for 72
73
ETMAAL 2013
BOOK OF ABSTRACTS
coping with the demands of everyday life and work in a complex, advanced society” (OECD, 2000, p. 13). Nevertheless, even the majority (65%) of the people with the lowest literacy has computer experience. A large proportion of them uses Internet (87%), and considers themselves to have sufficient computer related skills (72%) (Fouarge et al., 2011). This makes the Internet a suitable medium to communicate with low literates. In response to the somewhat worrying statistics concerning low and limited literacy, various interventions have been developed for patients with low literacy (Clement, Ibrahim, Crichton, Wolf, & Rowlands, 2009). The majority of those interventions however, lack solid background in theory (Wallace et al., 2012). Furthermore, several researchers addressed the importance of information accessibility by discussing readability (Hill-Briggs, Schumann, & Dike, 2012). For example, it is widely recommended that health materials should be written at 6th grade reading level maximally. Although often assumed, it is not ascertained that just simplifying health materials inevitably leads to better comprehension (Davis et al., 1998). Hardly any studies addressed message processing and message comprehension among people with low health literacy. This study therefore focuses on two factors that are expected to influence comprehension; text structure and language concreteness. Research showed that less skilled readers find it difficult to rely on text structure and to allocate attentional resources while reading (Rapp et al., 2007). Therefore, it can be expected that a message in which the gist is presented first will gain optimal attention of low literate readers. Consequently, this will be better processed and comprehended than when essential information is not presented at the beginning. In addition, based on Paivio’s dual coding theory, it can be expected that a message composed by means of concrete language improves message comprehension as concrete messages are easier to process than abstract messages (Paivio, 1986; Sadoski et al., 1993). Thus, the aim of this study is to investigate whether health messages tailored to reading skills and processing ease (by means of structure and concreteness respectively) improve message processing (recognition and recall) and comprehension of online textual health information among people with low health literacy. To test message effectiveness, four different health messages will be developed and pre-tested. Subsequently, the study will be executed by means of a 2 (low vs. high health literacy, measured) x 2 (structured vs. unstructured information, manipulated) x 2 (abstract vs. concrete information, manipulated) between subjects design. The study will be conducted within the next months. At the Etmaal van de Communicatiewetenschap 2013, the preliminary results of this study will be presented. This study provides insight in how tailored health messages are processed by individuals with low health literacy and how this affect comprehension of health materials. Furthermore, the results of this study can be used to adapt online health materials in order to improve comprehension among low literate audiences.
73
74
ETMAAL 2013
BOOK OF ABSTRACTS
MOOD AND TIME INFLUENCE DEFENSIVE AFFECT REGULATION FOLLOWING SELF-THREATENING HEALTH INFORMATION: EVIDENCE AT THE IMPLICIT LEVEL - ENNY DAS AND JONATHAN VAN 'T RIET Threatening health information aims to encourage and motivate people to engage in healthful and disease-preventive behaviors. However, many studies have documented that exposure to such messages can result in denial and dismissal of the information, especially when the information is personally relevant (Jemmot, Ditto & Croyle, 1986; Liberman & Chaiken, 1992). Recently, studies have shown that a positive mood can reduce defensive reactions to threatening health information (Das & Fennis, 2008; Das, Vonkeman, & Hartmann, 2012). The present research aimed to further investigate immediate and delayed effects of mood on affect regulation and defensive responses. Specifically, the objective was to assess the effects of positive and negative mood on short- and longer-term implicit affect regulation and persuasion following exposure to a threatening health message. We expected that exposure to threatening health information would hinder affect regulation of individuals at risk under negative mood conditions but not under positive mood conditions. We further expected that exposure to threatening health messages would induce most pronounced effects on self-regulation and persuasion immediately after message exposure. It was assumed that a positive mood and a time delay would provide individuals with the resources to deal with a difficult situation, and thus increase affect regulation and persuasion under high vulnerability conditions. Method Participants and design 174 participants were randomly assigned in a 2 (vulnerability: low vs. high) x 2 (mood: negative vs. positive) x 2 (timing of measurement: without or with delay) between participants design. Procedure and materials Vulnerability was successfully manipulated by providing participants with false feedback regarding toxicity levels of their personal hygiene products (low vulnerability: safe; high vulnerability: toxic). Mood was successfully manipulated following the procedure described by Raghunathan and Trope (2002). Participants all read the same threatening health message about toxic elements in personal hygiene products. Participants in the delayed measurement condition were given a 10-minute filler task. Participants in the no delay condition immediately proceeded to a Lexical Decision Task, which assessed implicit affect regulation based on response times to negative affect and positive affect words among 140 words, consisting of 70 existing words and 70 non-words (cf. Koole & Jostmann, 2004, Study 2; Koole & Fockenberg, 2011). Participants reported their attitude toward the threatening health message (1 item) and behavioral intentions to engage in safer use of personal hygiene products (3 items; Cronbach’s α = .87). Results Lexical Decision Task Mean RTs for positive and negative affect words were subjected to a 2x2x2x2 mixed ANOVA with 74
75
ETMAAL 2013
BOOK OF ABSTRACTS
repeated measures on word type. The analysis showed a main effect for word type, F(1, 164) = 50.52, p<0.001, a main effect for mood, F(1, 164) = 3.90, p<0.050, a significant interaction between mood, timing, and word type, F(1, 164) = 4,06, p<0.046, and a marginally significant four-way interaction between vulnerability, mood, timing, and word type, F(1, 164) = 3,20; p<0.075. Simple effects analyses revealed that positive mood speeded up reaction times to negative affect words (but not positive affect words) only under conditions of high vulnerability, and only at T1, i.e., without delayed measurement, F(1, 164) = 4.09; p=.045. No other simple effects were found. See Figures 1 and 2. Persuasion A 2x2x2ANOVA on message attitudes revealed a marginally significant main effect for vulnerability, F(1, 157) = 3.06, p=.085, and a marginally significant 3-way interaction between vulnerability, mood and timing, F(1, 157) = 3.32, p=.07. Timing of measurement only significantly affected attitudes toward the message under conditions of high vulnerability and negative mood, F(1, 157) = 4.82, p=.03. See Figure 3. A 2x2x2 ANOVA on intentions revealed a significant main effect for vulnerability, F(1, 157) = 30.20, p<.001. High vulnerability participants reported higher intentions to change their use of beauty products (M = 3.50, SD = 1.69) than low vulnerability participants (M = 2.23, SD = 1.19). Discussion Defensiveness was greatest for participants in the high vulnerability condition with a negative mood. These participants displayed a defensive mindset, with slowest responses to negative affect words and most negative attitudes towards the messages of all conditions. Slow response times to negative (but not positive) affect words suggest that individuals are keeping negative information at bay. This defensive mindset was only apparent at immediate assessment, and disappeared at a delayed assessment. There was no effect of mood and time of measurement on behavioral intentions. Intentions were only affected by vulnerability, with high vulnerability participants indicating stronger intentions than low vulnerability participants. Findings suggest that both a positive mood and a little time increase implicit affect-regulation and the acceptance of threatening health information.
Session #2 Thursday 7 February 15:15 – 16:15 hrs. Chair: Anika Batenburg Location: Praag (M3-05) Dutch session
COMMUNICATIE OVER GEWICHTIGE ZAKEN MET DE PRAKTIJKONDERSTEUNER IN DE HUISARTSENPRAKTIJK - SONJA VAN DILLEN, GERT JAN HIDDINK, JANNEKE NOORDMAN AND SANDRA VAN DULMEN Academisch vakgebied: Zowel Wageningen Universiteit als het NIVEL houden zich bezig met communicatie in de huisartsenpraktijk. Uit recent vragenlijstonderzoek onder huisartsen, uitgevoerd door Wageningen Universiteit, is gebleken, dat huisartsen eerder geneigd zijn te communiceren over beweging dan over voeding. De meerderheid vertoonde dezelfde communicatiepatronen wat betreft 75
76
ETMAAL 2013
BOOK OF ABSTRACTS
voeding en beweging. Vervolgens voerde Wageningen Universiteit een reviewstudie uit over de belangrijkste uitkomsten van onderzoek over communicatie over voeding en/of beweging tussen huisartsen en patiënten ter preventie of behandeling van overgewicht of obesitas. Geconcludeerd werd dat de inhoud van adviezen over voeding en beweging vrij algemeen is en dat een gecombineerd leefstijladvies zelden voorkomt. Observationeel onderzoek van het NIVEL wees uit, dat huisartsen voornamelijk over leefstijl spreken als dit gerelateerd is aan de klacht van de patiënt. Ook onderzocht het NIVEL hoe patiënten over leefstijl communiceren met praktijkondersteuners. Binnen de huisartsenpraktijk wordt de praktijkondersteuner steeds meer ingezet voor de zorg aan chronisch zieken. Praktijkondersteuners hebben veel kennis van specifieke aandoeningen en relatief meer tijd voor begeleiding, waardoor zij een belangrijke rol kunnen spelen in de ondersteuning van patiënten in de huisartsenpraktijk bij de dagelijkse omgang met hun ziekte. Onderwerp: Het aantal mensen met overgewicht of obesitas in Nederland is in dertig jaar tijd flink toegenomen. In de praktijk betekent dit dat huisartsen en praktijkondersteuners in toenemende mate op hun spreekuur patiënten met gewichtsproblematiek zien. In het huidige onderzoek staat de interactie tussen praktijkondersteuners en patiënten met overgewicht of obesitas centraal, in het bijzonder de frequentie, inhoud en kwaliteit hiervan. Hoe vaak communiceren praktijkondersteuners over gewicht, voeding en beweging in hun contacten met patiënten met overgewicht of obesitas? Wat is de inhoud van de adviezen over gewicht, voeding en beweging die gegeven worden door praktijkondersteuners aan patiënten met overgewicht en obesitas? Wat is de kwaliteit van deze adviezen over gewicht, voeding en beweging? Onderzoeksmethoden: In 2010 heeft het NIVEL video-opnamen gemaakt van consulten tussen praktijkondersteuners en patiënten. Hieruit werden 100 consulten van patiënten met overgewicht of obesitas geselecteerd. Geselecteerd werden consulten waarbij praktijkondersteuners overgewicht of obesitas registreerden als klacht van de patiënt, of waarbij praktijkondersteuners overgewicht/obesitas, gewicht, voeding of beweging/sport als klacht van de patiënt aanduidden in de ingevulde vragenlijst na het consult. Consulten werden alleen geïncludeerd als de praktijkondersteuner het gewicht van de patiënt bepaalde en als de patiënt ook daadwerkelijk overgewicht vertoonde. De 100 geselecteerde video-opnamen werden geobserveerd aan de hand van een ontwikkelde observatielijst. De observatielijst bestond onder andere uit items over het gewicht, frequentie van gespreksonderwerpen en initiatiefnemer daartoe, inhoud van adviezen en beoordeling als algemeen dan wel specifiek, gecombineerd leefstijladvies, en kwaliteit van adviezen. Als theoretisch kader werd hierbij het 5A’s Model gebruikt (Assess, Advise, Agree, Assist, Arrange). Naar aanleiding van een pilot werd de observatielijst aangepast. Ook werd de interbeoordelaarbetrouwbaarheid bepaald door 10 consulten te laten scoren door een tweede beoordelaar. Voorlopige resultaten: De dataverzameling werd onlangs afgerond. Momenteel is de verwerking en analyse van de resultaten gaande. Met behulp van inhoudsanalyse zullen de adviezen in verschillende categorieën worden ingedeeld. Vervolgens zullen overkoepelende thema’s worden geïdentificeerd. De voorlopige resultaten over de inhoud van de adviezen zullen gepresenteerd worden op het Etmaal van de Communicatiewetenschap.
76
77
ETMAAL 2013
BOOK OF ABSTRACTS
TRANSMURAL PALLIATIVE CARE BY MEANS OF TELECONSULTATION: A WINDOW OF OPPORTUNITIES …AND NEW RESTRICTIONS - JELLE VAN GURP, MARTINE VAN SELM, EVERT VAN LEEUWEN AND JEROEN HASSELAAR Background Audio-visual teleconsultation is expected to help home-based palliative patients, hospital-based palliative care professionals, and family physicians to jointly design better pro-active care. Consensual knowledge of the possibilities and limitations of teleconsultation in transmural palliative care is, however, largely lacking. The purpose of this study was to describe elements of the physical workplace and the cultural-social context of the palliative care practice, both of which are imperative to implementing a teleconsultation technology. We therefore analyzed palliative care professionals’ initial assumptions about the intermediate position of teleconsultation technology in care giving as well as their wishes about its design and function. These assumptions and wishes contain latent presumptions about the practice’s physical workplace and latent elements of the cultural-social context: real and relevant presuppositions that define the implementation of teleconsultation. In line with Don Ihde’s considerations [9, 10] about humans adopting new technologies, we think that teleconsultation technology is multistable: its eventual appearance and function, its fit, depends on whether and how participants of a particular practice, in this case the palliative care practice, adopt this technology in both their particular physical workplace (e.g., Internet connections and rooms for private conversations) and in their particular cultural-social context, i.e., their collective normative frameworks and daily routines [2, 10]. Methods A semi-structured expert meeting and qualitative, open interviews were deployed with professionals playing different roles in a university medical center based expertise team for palliative care. The aim was to explore professionals’ assumptions and wishes, which are considered to contain latent presumptions about the practice’s physical workplace and latent elements of the cultural-social context, regarding (1) the mediating potential of audio-visual teleconsultation, (2) how the audiovisual teleconsultations will affect medical practice, and (3) the design and usage of the teleconsultation technology. We used a qualitative analysis to investigate how palliative care professionals interpret the teleconsultation package in preparation. The analysis entailed open and axial coding techniques developed in a grounded theory approach. Results Respondents assume that 1. teleconsultation will hinder physical proximity, thereby compromising anamnesis and diagnosis of new or acutely ill patients as well as “real contact” with the person behind the patient; 2. teleconsultation will help patients becoming more of a pivotal figure in their own care trajectory; 3. they can use teleconsultation to keep a finger on the pulse; 4. teleconsultations have a healing effect of their own due to offered time and digital attention; 5. teleconsultation to open up an additional “gray” network outside the hierarchical structures of the established chain of transmural palliative 77
78
ETMAAL 2013
BOOK OF ABSTRACTS
care. This network could cause bypassing of caregivers and uncertainty about responsibilities; 6. teleconsultations lead to an extended flow of information which helps palliative care professionals to check the stories of patients and medical specialists. Conclusions Teleconsultation technology will facilitate experiencing through technology [9] for both patients and caregivers. This kind of mediation will extend certain bodily capacities and neutralize others, thereby magnifying certain aspects of the observed world and reducing others: respondents expect teleconsultation technology to extend hospital-based professionals’ perceptions of patients. Further, continuous, interpersonal, digital contact between patients and caregivers might magnify a patient’s healing capacity [15]. With regard to the magnification/reduction structure [9], Ihde claims that, normally, “what is revealed [by technology] excites, and what is concealed may be forgotten”. It is, however, not so much the magnification as the reduction that stands out for the interviewed palliative care professionals: for them, teleconsultation technology compromises connectedness with the patient in that it does not create physical proximity and diminishes professionals’ capacity to get in touch with the person behind the patient and develop a “gut feeling”. Besides being something through which caregivers get in touch with their patients, the teleconsultation technology is also an instrument palliative care professionals look at to interpret the patient’s status. Interpreting an image of the home is part of the palliative care physician’s anamnesis and has to be learned [9, 20]. The teleconsultation equipment itself also functions as a medium for transporting socio-cultural contexts. It is presumed to create a more pluricultural environment [9] in home and hospital. The hospital culture may intrude in the home through the technology and vice versa. These interactions of socio-cultural contexts might have consequences for the use of life-prolonging treatment, terminal hospital admissions, and multidimensional decision-making at life’s end, as it might result in giving patients a feeling of control over their culturally bound contacts with caregivers, and in a definite blending of different domains (home and hospital) into one still developing transmural area.
“IS ECHT BELANGRIJK NAAR DE HUISARTS TE GAAN ALS HET AANHOUDT. VANUIT ANESTHESIOLOGIE KUNNEN WE HIER NIETS MEE.” DE COMMUNICATIEVE UITDAGINGEN VAN TWITTERSPREEKUREN VOOR GEZONDHEIDSPROFESSIONALS - JOYCE LAMERICHS AND CHARLOTTE VAN HOOIJDONK “Be good and let others tell about it”. Zo omschrijft het HagaZiekenhuis in Den Haag de strategie van hun social media beleid. Ze bleken succesvol: in 2011 wonnen ze de verkiezing Beste Ziekenhuis 2.0. Een blik op de bijna 1500 tweets van het HagaZiekenhuis op Twitter leert ons dat er veel mededelingen worden gedaan over actuele zaken (een nieuwe wensboom op de afdeling oncologie; urologen en longartsen gaan op nascholing in Madurodam) en daarnaast wordt er ingegaan op praktische vragen en kwesties die bezoekers en patiënten inbrengen via Twitter. Die kwesties lopen uiteen van een vraag over een weigerende slagboom bij de ingang van het ziekenhuis tot 78
79
ETMAAL 2013
BOOK OF ABSTRACTS
complimenten van een patiënt voor de ontvangst bij het afspraakbureau. Opvallend is misschien dat het winnende ziekenhuis als het gaat om social media er in hun marketing en communicatiebeleid niet voor heeft gekozen om direct met patiënten in contact te treden via Twitterspreekuren. Een aantal Nederlandse ziekenhuizen doen dat echter wel. Op vooraf aangekondigde tijdstippen kunnen patiënten vragen stellen aan artsen over medische aandoeningen via Twitter. In onze presentatie doen we verslag van een onderzoek naar 38 Twitterspreekuren van 3 Nederlandse ziekenhuizen. Het gaat om het Rijnstate Ziekenhuis in Arnhem, het Jeroen Bosch Ziekenhuis in Den Bosch en het Maxima Medisch Centrum in Veldhoven-Eindhoven. Een van de aanleidingen voor deze studie is dat er weinig onderzoek is gedaan naar de manier waarop deze spreekuren daadwerkelijk verlopen. Wel is bekend dat de professionals het Twitterspreekuur niet altijd als gemakkelijk ervaren. Bijvoorbeeld omdat vragen niet altijd duidelijk zijn verwoord, maar ook omdat het beantwoorden van vragen in 140 tekens (het maximum voor een tweet) niet eenvoudig blijkt. Een eerste verkenning van het materiaal leerde ons dat er bovendien heel veel strategieën lijken te worden gehanteerd door de professionals die het Twitterspreekuur bemensen: van zeer zakelijk en feitelijk tot voorzichtig humor inzetten. Hoe worden die verschillende strategieën ontvangen? Zijn er aanwijzingen dat de ene strategie beter werkt dan de ander? Op deze vragen hebben wij met ons onderzoek een antwoord willen geven. In dit onderzoek is gebruik gemaakt van kwantitatieve en kwalitatieve onderzoeksmethoden om de interactie tijdens de Twitterspreekuren in detail in kaart te brengen. De tweets zijn verzameld via Twitter en vervolgens geanonimiseerd en gecodeerd. Door middel van statistische toetsen is de verdeling van relevante variabelen in kaart gebracht (bijvoorbeeld hashtags, vragen en antwoorden). Daarna is gebruik gemaakt van inzichten uit de conversatie analyse om opvallende aspecten uit de dataset verder te verkennen. Tijdens deze presentatie willen we rapporteren over twee opvallende en interessante bevindingen uit ons onderzoek. Dat is in de eerste plaats hoe patiënten vragen stellen via Twitter en hoe hierop wordt gereageerd door de artsen. In tegenstelling tot een verkenning door twee Nederlandse huisartsen uitgevoerd in 2010, die stelden dat patiënten en huisartsen gemiddeld 4 tweets nodig hebben om een vraag/antwoord paar af te sluiten, lijkt dat cijfer lager te liggen in de Twitterspreekuren met artsen en specialisten verbonden aan ziekenhuizen. Wat we daarnaast zien in de antwoordformulering van de artsen is dat zij in die formulering niet alleen rekening houden met de vraag van de vragensteller, maar tegelijkertijd hun taalgebruik afstemmen op het brede publiek dat de tweets raadpleegt. Bijvoorbeeld doordat zij in hun reactie, medische terminologie ´vertalen´ naar een gebruikelijkere term (´botontkalking` in plaats van ´osteoporose´). Op de tweede plaats bleken hashtags (aangegeven door middel van het #-teken) op interessante manieren te worden ingezet in tweets aan het Twitterspreekuur. Hashtags lijken voor de deelnemers aan het Twitterspreekuur niet alleen een organisatorische functie te vervullen (de praktijk van ‘tagging’ door gebruikers van Twitter om tweets terug te kunnen zoeken), maar dienen ook een interactioneel en conversationeel doel. Vooral patiënten bedienen zich van dit laatste type hashtags en voegen daarmee een evaluatieve lading toe aan hun tweet. Zo lijken sommige hashtags (bijvoorbeeld #verslaafdaan) gebruikt te worden om een respons bij de arts uit te lokken over een medische kwestie. Tijdens onze presentatie zullen we verder ingaan op welke evaluatieve lading er 79
80
ETMAAL 2013
BOOK OF ABSTRACTS
met uiteenlopende hashtags precies wordt toegevoegd aan een tweet, en hoe hierop wordt gereageerd door de arts. Ook laten we zien hoe patiënten en artsen op verschillende manieren gebruik maken van dit type hashtags. We zullen in onze presentatie ook ingaan op aanbevelingen voor gezondheidsprofessionals die ingegeven worden door de resultaten uit dit onderzoek en die specifiek gericht zullen zijn op het hanteren van typen hashtags in tweets en het formuleren van antwoorden door professionals.
KENNIS OVER CHEMOTHERAPIE: EEN EXPLOREREND ONDERZOEK NAAR FACTOREN DIE VAN INVLOED ZIJN OP KENNIS VAN PATIËNTEN EN NIET-PATIËNTEN - JULIA VAN WEERT AND SIFRA BOLLE Achtergrond en doel In veel gezondheidscommunicatie theorieën wordt kennis als een belangrijke voorwaarde beschouwd voor gedrag, zoals bijvoorbeeld het opvolgen van adviezen en zelfmanagement (Maibach & Cotton, 1995). Eerder onderzoek heeft uitgewezen dat kennis van patiënten een positieve relatie heeft met therapietrouw, patiënttevredenheid en vermindering van stress gerelateerde symptomen (Bishop, Barlow, Hartley & William, 2003; Cameron, 1996; Galloway et al., 1997; Kessels, 2003; Mossman, Boudioni & Slevin, 1999). Chemotherapie is één van de drie meest gebruikte behandelingen bij kanker (bron: www.kwf.nl). Goede voorlichting dient zo goed mogelijk aan te sluiten op kennis die (nieuwe) patiënten al dan niet reeds hebben. Het is daarom belangrijk om niet alleen inzicht te hebben in de kennis van chemotherapiepatiënten, maar ook van nietchemotherapiepatiënten. Publieke kennis van chemotherapie is echter nog maar weinig onderzocht. Deze studie heeft in de eerste plaats tot doel om het kennisniveau van chemotherapiepatiënten en niet-chemotherapiepatiënten te onderzoeken. Dit geeft enerzijds inzicht in de mogelijke hiaten in kennis van chemotherapiepatiënten, en anderzijds in het gemiddelde kennisniveau dat mensen die niet eerder behandeld zijn met chemotherapie hebben bij de start van de behandeling. In de tweede plaats heeft de studie tot doel om factoren die van invloed zijn op het kennisniveau vast te stellen. Voor dit doel wordt een model getoetst van factoren waarvan op basis van theorie en eerder onderzoek een mogelijke invloed op kennis verondersteld kan worden. Methode In dit onderzoek is gebruik gemaakt van een online vragenlijst, die door 2785 respondenten volledig is ingevuld. Het meetinstrument voor het meten van kennis is speciaal ontwikkeld voor deze studie en bestaat uit 33 items (range 0-33). Bij de analyses over kennis is onderscheid gemaakt tussen 1) chemotherapiepatiënten (n = 999); 2) niet-chemotherapiepatiënten die zeer betrokken zijn (geweest) bij een naaste die chemotherapie kreeg (n = 821) en 3) niet-chemotherapiepatiënten die niet zeer betrokken zijn geweest bij een naaste met chemotherapie (n = 1162). Naast verschillen tussen chemotherapiepatiënten en niet-chemotherapiepatiënten is de invloed van sociodemografische en ziektespecifieke achtergrondkenmerken (‘leeftijd’, ‘geslacht’, ‘opleidingsniveau’, ‘kwaliteit van leven’, ‘tijd sinds diagnose’ en ‘curatief of palliatief behandeldoel’), psychologische factoren (‘beliefs ten aanzien van kanker’, ‘monitoring coping stijl’ en ‘stressreacties’), onvervulde informatie- en communicatiebehoeften en het gebruik van informatiebronnen (‘huisarts’, ‘alternatieve genezer’, ‘patiëntenvereniging’, ‘andere patiënten face-to-face’, ‘andere patiënten via Internet’, ‘brochures’, ‘Internet’ en ‘boeken’) op kennis onderzocht. 80
81
ETMAAL 2013
BOOK OF ABSTRACTS
Resultaten Chemotherapiepatiënten hebben meer kennis van chemotherapie dan niet-chemotherapiepatiënten, behalve wanneer een niet-chemotherapiepatiënt zeer betrokken is (geweest) bij een chemotherapiepatiënt. Chemotherapiepatiënten met meer onvervulde informatie- en communicatiebehoeften hebben minder kennis van chemotherapie dan chemotherapiepatiënten die wel tevreden zijn over de mate waarin aan hun informatie- en communicatiebehoeften is voldaan door de behandelend artsen en/of verpleegkundigen. Wat betreft de psychologische factoren zijn de beliefs, ofwel overtuigingen die men heeft over kanker, van invloed. Chemotherapiepatiënten met sterkere negatieve overtuigingen ten aanzien van kanker hebben minder kennis van chemotherapie dan mensen die minder negatief tegen de ziekte kanker aankijken. Daarnaast hebben patiënten met een monitoring, ofwel informatiezoekende, copingstijl meer kennis dan mensen die minder informatiezoekend zijn ingesteld. Mensen die veel kanker-gerelateerde stressreacties ervaren hebben eveneens meer kennis. Ook het gebruik van aanvullende informatiebronnen, met name een patiëntenvereniging, draagt bij aan de kennis. Tot slot zijn er een aantal socio-demografische kenmerken van invloed op kennis. Dit geldt voor zowel chemotherapiepatiënten als nietchemotherapiepatiënten. Naarmate men ouder is, heeft men minder kennis van chemotherapie, vrouwen hebben meer kennis van chemotherapie dan mannen en middelhoog en hoog opgeleiden hebben meer kennis van chemotherapie dan laag opgeleiden. Niet-chemotherapie patiënten die zeer betrokken zijn (geweest) bij iemand met chemotherapie hebben meer kennis dan nietchemotherapie patiënten die dat niet zijn. Conclusie Het afstemmen van de informatie en communicatie op de behoeften van de patiënt is een voorwaarde voor goede voorlichting over chemotherapie en leidt tot meer kennis. De resultaten van dit onderzoek geven handvatten om de voorlichting aan patiënten over een behandeling met chemotherapie optimaal aan te laten sluiten op persoonlijke kenmerken en bestaande kennis.
Session #3 Thursday 7 February 16:45 – 17:45 hrs. Chair: Enny Das Location: Praag (M3-05) English/Dutch session
SWINE FLU IN THE E.R.: THE RELATIONSHIP BETWEEN MEDIA ATTENTION FOR THE SWINE FLU AND REGISTRATION IN AN EMERGENCY MEDICINE UNIT - SARA NELISSEN, KATHLEEN BEULLENS, MARC SABBE AND JAN VAN DEN BULCK When the world is faced with a new threatening pandemic outbreak, fear and panic can occur on a large scale (Bonneux & Van Damme, 2006). Because of the high death rate of earlier pandemic outbreaks, new forms of influenza remain a considerable cause for concern for humanity (Del Rio, 2009). Furthermore, the media report heavily about these events. This high coverage could lead to even higher levels of worry and fear among the population (Sandman, 2002). Previous research (Van 81
82
ETMAAL 2013
BOOK OF ABSTRACTS
den Bulck & Custers, 2009) on the Avian flu outbreak of 2005 showed that exposure to television was highly related to worrying about this illness. In 2009 a new form of the influenza virus (A/H1N1) emerged in Mexico. This new disease knew a fast spreading rate (WHO, 2009) and was called “the Swine Flu”. Content analytical research showed that the Swine Flu received extensive coverage in newspapers worldwide (Duncan, 2009; Hilton & Hunt, 2010). These observations lead to the question whether such reporting’s may affect people’s behavior and may therefore aggravate or alleviate the burden of a pandemic (Sandman, 2000). The central question of this study is whether news of the Swine Flu reflects the course of the disease (when more people get ill, more attention is devoted to news about the Swine Flu) or whether reporting affects patient behavior (more people present themselves with symptoms as a result of exposure to worrying news reports). This article looks at the number of people who presented themselves to the Emergency Room (E.R.) of a large teaching hospital in Flanders, assuming they displayed symptoms of the Swine Flu and relates it to the frequency of articles in the Flemish news press for the same time span. RQ1 Is there a relationship between the number of self-presentations to the E.R., and the number of articles concerning the Swine Flu, published in Flemish newspapers? RQ2 Does an increase in newspaper reports of the Swine Flu follow an increase in self-presentations (news as mirror of the disease) or vice-versa (E.R. visits as mirror of news attention)? RQ3 Is the relationship between self-presentations at the E.R. and news attention different for quality newspapers than for tabloid newspapers. The (Hospital name deleted for review) is one of the most respectable hospitals in Flanders. During the pandemic, this hospital designed a special Emergency Unit especially for potential Swine Flu cases. They collected data from August 2009 until January 2010 of all persons that came to the unit, thinking they had the flu and of all the patients who were actually diagnosed with Swine Flu. We compared this number of potential patients with the number of articles concerning the Swine Flu that were published in seven Flemish newspapers. Via the press database Mediargus we generated an overview for the selected time frame. To examine the relationship between these two time series we looked at the cross-correlation function in SPSS to determine the direction of this relationship. “The cross-correlation function indicates a direct or delayed dependence between two different time series” (Box & Jenkins, 1970). We found a significant high correlation coefficient between the two time series (r=.728, p<.01). Furthermore, we wanted to find out whether there was a difference between articles of quality papers or tabloid papers. We found a significant correlation (p<.01) between the number of selfpresentations and the number of quality articles (r=.700) and of tabloid articles (r=.714). These results suggest that the number of people who thought they had the Swine Flu and the number of articles in the news press were closely related and that there was no difference in this relationship between quality and tabloid newspapers. Finally, we found that the number of patients was the leading indicator. Thus, whenever the number of potential patients increased, the press wrote more about the Swine Flu. Our results indicate that in this case, the media reflected reality. Increased attention in the media 82
83
ETMAAL 2013
BOOK OF ABSTRACTS
followed increases in the number of patients. Earlier research suggested a reverse relationship (e.g. Van den Bulck & Custers, 2009). So far, little is known about the impact of news on the anxiety levels of the public. This study tried to increase our understanding of the effects of news reports, during disease outbreaks, on fear for that diseases. Further research in this area is warranted.
SUFFERING AS A DISCIPLINE? SCHOLARLY ACCOUNTS ON THE CURRENT AND FUTURE STATE OF RESEARCH ON MEDIA AND SUFFERING - STIJN JOYE An emerging field of research within humanities and social sciences concerns itself with the issue of suffering. Following its growing (mediated) societal prevalence and impact in recent years, forms of social suffering have intrigued an increasing number of scholars and have already spurred a rich body of work, reflecting an astonishing richness in terms of approaches, theoretical perspectives and topics of study. Many have recognized the research on media and suffering as a legitimate, timely and necessary field of study, but alongside its emergence within academia, questions arose on its disciplinary home, scope and nature. The purpose of this paper is twofold. Firstly, we address the key question of (inter)disciplinarity by charting the different strands of social science studies on suffering, with a particular focus on the relation between (news) media and suffering. Secondly, we aim to identify future directions to move the research forward. In order to do so we draw on a literature review and semi-structured elite interviews with twelve leading scholars in the field: Jonathan Benthall, David Campbell, Lilie Chouliaraki, Simon Cottle, Suzanne Franks, Paul Frosh, Folker Hanusch, Susan Moeller, Shani Orgad, Mervi Pantti, Keith Tester and Karin Wahl-Jorgensen. Transcriptions of the interview data were analyzed by a theme-driven qualitative content analysis. Emerging from the data was a strong reluctance of the twelve interviewees towards conceiving the ongoing research on media and suffering as a discipline on its own. All interviewees preferred to look at it as a constellation of people, working from different perspectives and backgrounds but all interested in the same topic. In other words, an area of research that is defined by an inter- and multidisciplinary coming together of research interests, expertise and people. The inherent social nature of suffering and its status as a fundamentally human experience informed this debate and resulted in a positioning of the research on media and suffering at the heart of social sciences and humanities as well as at the crossroads of different disciplines. This overall open view was also reflected in the discussion on the scope of the research and the future directions in research. The scholars identified a wide range of driving concepts and key issues which we can summarize by the following tropes: mediation, emotions, audiences and societal trends. However, the interviewees tended to stress the need for some measure of conceptual hygiene, a potential drawback of the interdisciplinary nature of the field. Regarding future directions, the scholars pointed towards a number of directions to move the research forward. Especially empirical audience research is high on the academic agenda as are studies that look into the role of new media with regard to (witnessing) suffering. Other widely shared comments included a further opening up of the research in terms of methodological and disciplinary approaches. To conclude, this paper has the purpose of initiating further interdisciplinary dialogue and debate
83
84
ETMAAL 2013
BOOK OF ABSTRACTS
within social science and humanities to come to a more comprehensive understanding of suffering and its (mediated) experience.
E-MAIL VERSUS CHAT COUNSELING: A COMPARISON OF PROBLEM PRESENTATIONS - WYKE STOMMEL AND FLEUR VAN DER HOUWEN In June 2012, the Dutch minister of Public Health launched a permanent e-health monitor with the aim of encouraging health care organizations to use e-health.1 In the domain of health promotion, online counseling has been found effective in the sense that clients report improvements with regard to their problems (Riper et al. 2007). Research on actual interaction between clients and coaches in online counseling is still scarce, but increasing. It appears that the interactional dynamics we find in spoken counseling also play a role in online counseling. For example, requests have been found to be designed in such a way that the counseling relationship is not jeopardized (Harris et al. 2011). Also, formulations in online counseling are similarly used for “therapeutic” actions (Stommel and Van der Houwen forthc.). Counseling through an online modality, however, also appears to have medium specific characteristics. For instance, Danby et al. (2009) have shown that certain actions in chat, namely formulations following problem presentations, are more ambiguous than the equivalent actions on the phone. It seems, thus, that various kinds of online counseling (e.g., chat counseling, e-mail counseling) impact the unfolding interaction in their own way. Some differences seem evident: e-mail counseling messages are probably longer than chat posts, e-mail counseling is probably not restricted to time while chat counseling is, etc. However, the implications for counseling are unclear. To our knowledge, there has been no research yet comparing various modalities of online counseling, such as chat versus e-mail counseling. In this paper, we present an explorative analysis of problem presentations in e-mail and chat counseling. The term “problem presentation” is used in studies of medical encounters to describe the phase right after the opening of the consultation in which patients tells about the problem(s) that brought them to their doctor. Problem presentations have been claimed to be important because when they are “accurate”, patients more frequently report a reduction of problems and are more satisfied about the consultation (Roter and Hall 2006). Frequently, the physician prompts the problem presentation with a question such as “What can I do for you today?”. When the presentation appears complete, the physician may close it by producing the first “history taking”question. Because e-mail does not have a question-answer structure that organizes the interaction (as chat), the closing of the problem presentation in e-mail is not sequentially related to the first history-taking question. Therefore, we analyze more broadly the interactional space in which the client first presents his/her problem(s). Our paper is positioned in the field of discourse analysis. We present a conversation analysis (CA) of e-mail and chat counseling. CA is a method which is increasingly used in the area of online interaction and our paper aims at contributing to this developing field. Specifically, we compare interactions that are differently mediated but institutionally identical (both are counseling interactions, have the same kind of therapeutic base line and the counselors have received the same training). Our data were collected as part of a research project on interactional aspects of online counseling. 84
85
ETMAAL 2013
BOOK OF ABSTRACTS
The counseling program was developed by a Department of Clinical Psychology for research purposes. It was offered to persons with moderate symptoms of anxiety and/or depression. Both the e-mail and the chat counseling encompassed 5 weeks. E-mail clients were invited to write one e-mail per week to their coach on which they received “feedback” and chat clients were invited to do one 25 minute chat session per week with a counselor. The materials we used for this study consist of 22 threads of e-mail counseling and 15 logs of clients’ first chat session with a counselor. Because of our interest in initial problem presentations, we selected the first two messages of each e-mail thread (consisting of the first introductory coach email and the first client e-mail). Preliminary results of our analysis include: - In e-mail counseling, counselors open up a broad interactional space to discuss various issues; in chat, counselors restrict problem presentations by giving the client less space to elaborate (orienting to one issue, rather than eliciting a narrative or a set of problems). - In e-mail counseling, clients use long narratives as problem presentations; in chat counseling, clients tend to construct problem presentations using a symptom (tiredness, grief) or a diagnosis (moderate depression). - In e-mail counseling, clients close their problem presentations referring to completeness; in chat counseling, counselors treat client’s problem presentations as incomplete. Our results shed light on what the implications are for counseling and we make preliminary statements on the affordances of chat versus e-mail for counseling purposes. 1 see: http://www.rijksoverheid.nl/nieuws/2012/06/07/minister-schippers-start-permanente-ehealth-monitor.html
COST-EFFECTIVENESS OF ADHERENCE ENHANCING INTERVENTIONS: A SYSTEMATIC REVIEW EDWIN OBERJÉ, MARIJN DE BRUIN AND SILVIA EVERS Objective Approximately 30% to 50% of patients with chronic illnesses do not take their medication as prescribed. There are good reasons for trying to enhance treatment adherence to prescribed drugs, because non-adherence not only limits the potential benefits of drug therapy, it also has substantial clinical and economic impact on a range of chronic diseases. In the US, cost estimates of nonadherence range between $100 and $300 billion annually in unnecessary healthcare spending and loss of productivity. Policy makers who decide about the implementation of health care services need to be able to compare the relative costs and benefits of available interventions. In the light of the acclaimed costs of non-adherence and the pressure to reduce health care expenditures in the current economic climate, a systematic review that assesses experimental evidence for costeffectiveness of adherence enhancing interventions would be timely. Methods MEDLINE, Econlit, PsycInfo, and NHS EED were searched for RCTs of economic evaluations of adherence enhancing interventions. Economic evaluations were included of interventions primarily aimed at enhancing medication adherence. Cost-effectiveness was determined by assessing the 85
86
ETMAAL 2013
BOOK OF ABSTRACTS
effects of interventions on adherence, clinical outcomes, quality of life, and costs consequences. The methodological quality was evaluated with the Drummond checklist and the risk of bias with the Cochrane bias assessment tool. Studies were categorised into three levels of evidence, ranging from best available evidence to studies that produce limited evidence. Results Eight studies assessed adherence data. Four of these eight studies showed a significant increase in adherence for the intervention group. Eleven studies reported intervention effects on at least one intermediate outcome, of which three studies showed significant results. None of the studies that produce best available evidence showed significant effects on adherence. Only one study that produced high quality evidence showed a significant improvement on intermediate outcomes. Other intervention effects were found in studies that produced limited evidence. Studies that did not use the societal perspective, scored remarkably lower on methodological quality, indicating that results from those studies should be interpreted with caution. Six studies used health-related or general quality of life as outcome variable, but none of them showed significant effects. Some studies were underpowered and larger sample sizes may be required to detect small but relevant significant differences in quality of life. However, it is not clear whether the non-significant quality of life outcomes are due to lack of power, a lack of effective interventions, insensitive outcome measures, short time horizons, other reasons, or a combination of these. Conclusions Despite that non-adherence is considered a major challenge for the cost-effectiveness of health care for chronic conditions, and many adherence studies have been conducted, only a limited number of RCTs is available that evaluate the cost-effectiveness of adherence enhancing interventions. Many of the studies show a poor quality of reported data and differ widely in terms of the assessed costs and consequences. It should be noted though that all high quality economic evaluations were published recently, suggesting that quality has improved over time. The conclusions concerning costeffectiveness were generally favourable for the medication adherence enhancing interventions, but none of the interventions showed significant improvements on quality of life. More significant improvements on adherence and intermediate outcomes were seen in studies that produce limited evidence.
86
87
ETMAAL 2013
BOOK OF ABSTRACTS
Session #4 Friday 8 February 09:00 – 10:00 hrs. Chair: Jan Van den Bulck Location: Praag (M3-05) English session
WHAT DO PRENATAL AND POSTNATAL WOMEN TALK TO THEIR HEALTHCARE PROVIDERS: FINDINGS FROM THE LOS ANGELES MOMMY AND BABY (LAMB) SURVEY - YUPING MAO AND LU SHI In the United States, Chinese immigrants have been the fastest growing ethnic minority population over the past two decades, and are the second largest immigrant group in the U.S. after Mexicans (Camarota, 2007). Chinese women’s health issues become an increasingly important social issue as this population grows. In the U.S. society, the Chinese are perceived as “the model minority” with higher income and educational level than the U.S. national average. However, in the U.S. context Chinese women’s health communication might not be as effective as their Caucasian counterparts. Researchers found that the loss of extended social networks and language barriers are the two primary reasons that Chinese immigrants have difficulties in accessing health information and called for research to understand Chinese women’s experiences of communication for health information (Ahmad, Shik, Vanza, Cheung, George, & Stewart, 2004). Through analyzing the Los Angeles Mommy and Baby (LAMB) 2007 survey, this research attempts to explore the content of doctor-patient communication during those women’s prenatal and postnatal physician visits. We also discuss the similarities and differences between Chinese women and the Caucasian women in their information exchange with their healthcare providers. The LAMB survey is a population-based survey with multi-level clustered sampling. The survey was developed by the Maternal, Child, and Adolescent Health Programs of Los Angeles County. The purpose of the survey was to examine maternal and child health indicators such as prenatal care, health behaviour during pregnancy, and postnatal recovery. The LAMB surveys were mailed to new mothers who had a live birth in the Los Angeles County within the past three months. To reach participants from different cultures, the surveys were conducted in different languages including English, Spanish, and Chinese. Tailored cover letters, incentives, multiple mailings and nonrespondent follow-up were applied to increase the response rate. In total, 4,518 of the 12,675 women reached by the multilevel sampling completed the survey, which results in a response rate of 35.6%. In our analysis, we focus on the results of 113 Chinese women and 376 Caucasian women. Using SAS 9.0’s PROC FREQ function, we analyzed the survey items on those women’s communication with healthcare providers at their prenatal and postnatal stages, including whether and why they meet healthcare providers for maternal issues, and the key information exchanged in their communication with healthcare providers. Part of the results confirmed the general perception of Chinese as “the model minority,” since Chinese women did a good job of seeking medical information at their prenatal stage: In their communication with healthcare providers, Chinese women are more likely to discuss medical and physical health issues such as healthy weight (62.5% vs. 57.5%) and blood pressure than Caucasian women (40.7% vs. 34.5%). However, Chinese women 87
88
ETMAAL 2013
BOOK OF ABSTRACTS
are significantly less likely to discuss health-related social and psychological issues such as domestic violence (0.0% vs. 18.9%) and depression with their healthcare providers (14.8% vs 27.7%) than Caucasian women. Chinese women’s reluctance in discussing social and psychological issues with healthcare providers is found at their postnatal stage as well. Chinese women are significantly less likely to discuss issues such as anxiety (11.5% vs 24.7%), depression (15.6% vs. 42.3%), and domestic/child abuse (5.5% vs. 11.7%) than Caucasian women in their postnatal visit to healthcare providers. The survey reveals that an important communication gap of Chinese women during their prenatal and postnatal stages is their communicative effort of seeking social and psychological help from healthcare providers. Maternity is exciting but also stressful for both women and their families. Although Chinese culture is a collectivist culture and American healthcare providers might assume that Chinese people tend to have a strong association with their communities, overseas Chinese are far away from and disconnected with their social networks in China. Compared with Caucasian American women, Chinese women in the U.S. may have greater needs of social and psychological support during maternity. In Chinese culture, psychological issues are considered as very personal issues and people tend to rely more on family members and friends than medical systems, which might lead to overseas Chinese women’s avoidance or ignorance of discussing those issues with medical professionals. This study suggests the current healthcare system should be aware of the group differences between Chinese women and Caucasian women and design culturally sensitive programs to provide social and psychological help to Chinese women during their prenatal and postnatal stages.
PUT YOURSELF IN THE PATIENT’S SHOES: THE EFFECT OF MODALITY AND COMMUNICATION STYLE ON WEBSITE SATISFACTION AND RECALL OF ONLINE HEALTH INFORMATION - NADINE BOL, ELLEN M.A. SMETS, HANNEKE C.J.M. DE HAES, EUGÈNE F. LOOS AND JULIA C.M. VAN WEERT Introduction Increasingly more health information is presented online and sometimes even exclusively online (Morrell, 2002). Due to the use of complex medical language and the additional barrier of presenting this complex information online, many people struggle with health information they find online. Especially older adults experience difficulties because of declines in basic cognitive functioning. To address these age-related issues and thereby increasing website satisfaction and recall health information among older adults, multimedia learning environments (e.g., health websites) that support older adults’ declining cognitive functions might be used for older adults (Paas, Van Gerven, & Tabbers, 2005). Such compensatory multimedia strategies include audiovisual information. In addition to message modality, different communication styles, such as informational (i.e., factual information, professional opinions) versus narrative (i.e., personal stories, individual experiences) styles have been found to account for variance in the effectiveness (i.e., website satisfaction and information recall) of health messages as well (Hinyard & Kreuter, 2007). It is often reported that especially narrative audiovisual information (vs. informational audiovisual information) enhances narrative effects, such as website satisfaction and information recall, due to higher involvement with the character and the story (identification and transportation). However, there is still no consensus about which modality and communication style is more effective to increase website satisfaction and information recall in older adults. Therefore, this paper will examine the effects of modality and 88
89
ETMAAL 2013
BOOK OF ABSTRACTS
communication style (written vs. informational audiovisual vs. narrative audiovisual information) on website satisfaction and recall of health information among younger and older adults. Method Younger (n = 195, mean age = 41.55 years) and older (n = 162, mean age = 68.92 years) adults participated in a 3 (written vs. informational audiovisual vs. narrative audiovisual information) x 2 (younger vs. older age) between-subjects factorial designed experiment. Participants were asked to fill out an online questionnaire which included one of the experimental stimuli. This was either a written version of a webpage including textual information on Radio Frequency Ablation (RFA) treatment, an informational video in which a physician explained RFA treatment, or a narrative video in which a patient explained RFA treatment. Information on RFA treatment was chosen because of its relative unknown character, which enhances the reliability of the recall measurement. Measures included satisfaction with the comprehensibility, attractiveness, and emotional support of the website, recall of information, identification, transportation, and background variables, such as age, education, and internet experience. Two separate ANOVAs were conducted to test the effects of modality and communication style on website satisfaction and information recall. Additionally, the conditional process modeling program PROCESS was used to test for the indirect effects of communication style through identification and transportation. Results Audiovisual information (vs. written information) was found to significantly increase recall of information, but only for the narrative video. No modality nor communication style effect was found for website satisfaction, indicating that message modality and communication style do not influence satisfaction with the comprehensibility, attractiveness, and emotional support of the website. However, regardless of modality and communication style, older adults reported to be more satisfied with the attractiveness and emotional support of the website than younger adults. Furthermore, the results showed that identification and transportation were important predictors of website satisfaction and recall of information. Identification was positively related to satisfaction with emotional support and negatively related to recall of information, indicating that identifying with the narrator in the story leads to enhanced satisfaction, but decreases recall of information. Transportation was found to be positively related to all outcome variables, showing that being merged into the narrative world leads to enhanced website satisfaction and better information recall. PROCESS showed marginally significant moderated mediation effects in the relationship between communication style and satisfaction with attractiveness, satisfaction with emotional support, and recall of information (at 90% CI). Identification was found to mediate this effect, but only in younger adults. We found that that narrative communication increased website satisfaction in younger adults through enhanced identification with the character, but at the same time decreased recall of information. Conclusions Both modality and communication style play an important role in explaining recall of health information. Showing narrative audiovisual information significantly increased recall of health information compared to written information. Identification and transportation were found to be important predictors of website satisfaction and information recall as well. Furthermore, identification marginally significantly mediated the relationship between communication style and 89
90
ETMAAL 2013
BOOK OF ABSTRACTS
our outcomes in younger adults. It was shown that identification leads to enhanced website satisfaction in younger adults, but at the same time leads to decreased recall of information in younger adults.
COMMUNICATION ABOUT CHEMOTHERAPY TREATMENT DECISIONS: TAKING PERCEPTIONS ABOUT TREATMENT CONSEQUENCES INTO ACCOUNT - SIFRA BOLLE AND JULIA VAN WEERT Background Nowadays, shared decision making is a frequently used term in research addressing provider patient communication. Although there are many different conceptualizations of shared decision making, patients’ values and preferences are often used concepts to define shared decision making (Makoul & Clayman, 2006). Patients can base their treatment preferences on perceptions about the consequences of a treatment on their quality of life. (Kiebert et al., 1994). Literature indicates that perceptions about the side effects of chemotherapy that patients have nowadays can be different from patients’ perceptions ten years ago. This can be explained by improved chemotherapy treatment and the improved treatment of side effects (Carelle et al., 2002). This study gives insight into the present perceptions about the side effects of chemotherapy. Furthermore, cancer patients' and non-cancer patients' perceptions about chemotherapy are likely to differ (Lindley et al., 1999). To our knowledge, little research has addressed the differences in perceptions between patients and non-patients. Insight in these differences gives us valuable information to improve shared decision making with cancer patients that will be treated with chemotherapy for the first time. Method An online survey (n=2890) was carried out to assess chemotherapy patients' and non-cancer patients’ perceptions about the side effects of chemotherapy and treatment decisions. Respondents were assigned to three groups: (1) (former) cancer patients that are or have been treated with chemotherapy (n = 999), (2) non-cancer patients that are highly involved with a significant other that has been treated with chemotherapy (n = 729), and (3) non-cancer patients that are not or not highly involved with a significant other that has been treated with chemotherapy (n = 1162). Perceptions of chemotherapy side effects were measured using fifteen possible side effects of chemotherapy. These side effects were selected on basis of an information leaflet for chemotherapy patients, that is published by the Dutch Cancer Society (KWF). The mentioned side effects were then revised by a panel of healthcare providers in oncology and reduced to a list of fifteen possible side effects. Chemotherapy patients were first asked to indicate which of the fifteen possible side effects they had actually experienced. Then the respondents were asked to indicate how much the experienced side effects bothered them on a four-point Likert-scale from ‘not at all’ to ‘very much’. Non-cancer patients were asked to indicate how much each of the fifteen possible side effects would bother them on the same four-point Likert-scale. Results Nine out of the fifteen measured side effects were perceived as significantly more bothersome by both groups of non-cancer patients than actually experienced by chemotherapy patients. This result was found for the side effects ‘anxiety and depression’, ‘bleeding’, ‘concentration and short term memory changes’, ‘nausea and vomiting’, ‘mouth and throat sores’, ‘infection and fever’, ‘nerve and muscle effects’, ‘constipation and diarrhea’, and ‘hair loss’. Four out of the fifteen measured side 90
91
ETMAAL 2013
BOOK OF ABSTRACTS
effects were perceived as significant less bothersome by both groups of non-cancer patients than actually experienced by chemotherapy patients. This result was found for the side effects ‘fatigue’, ‘changes in sexuality’, ‘hand-foot syndrome’, and ‘effects on fertility’. For the side effects ‘allergic reactions’ and ‘skin and nail changes’, no significant differences in perceptions between chemotherapy patients and non-cancer patients were found. The results for treatment decisions also showed a significant difference between chemotherapy patients and non-cancer that were not or not highly involved with a significant other that has been treated with chemotherapy: where 64.5% of chemotherapy patients were willing to accept an intensive treatment under any circumstances, only 56.4% of non-cancer patients indicated to be willing to accept the same intensive treatment under any circumstances. Conclusion Chemotherapy patients experience most side effects as less bothersome than non-cancer patients expect. Non-cancer patients that are highly involved with a significant other that has been treated with chemotherapy have the same perceptions about most chemotherapy side effects as non-cancer patients that are not or not highly involved with a significant other that has been treated with chemotherapy. Chemotherapy patients are more willing to accept an intensive chemotherapy treatment under any circumstances than non-cancer patients that are not or not highly involved with a significant other that is treated with chemotherapy. Practical implications are discussed and implications for future research are made. The results of this study can be used to improve provider patient communication.
SEWING THE SEEDS OF MOTHERHOOD: THE SOCIO-POLITICAL DYNAMICS OF SEWING MUMMY BLOGS - JISKA ENGELBERT AND AN KUPPENS The past years have witnessed an increasing proliferation and popularity of “sewing mommy blogs”. On these blogs, which can be regarded as a fusion between mommy blogs, craft blogs and fashion blogs, women write (mainly) about the clothes they design and/or make for their children and themselves, illustrating their posts with (generally highly stylized) pictures of their creations. While the advent of sewing mommy blogs appears to be a more or less global phenomenon, the FlemishDutch community presents an especially interesting case. For recently, the unexpected success of one particular sewing blogger’s book (“Allemaal Rokjes” by the Flemish Kim Leysen has sold over 50.000 copies across Flanders and the Netherlands) has lifted the current sewing hype out of the confines of the blogosphere. Moreover, the current trend for (highly educated) mothers to retreat from the labor market in order to focus on domestic activities which are typically regarded and condemned as traditional and disempowering (such as sewing, cooking, and raising their kids) has encountered criticism and become the subject of public debate. Flemish labor secretary Monica De Coninck, for instance, has warned against the risks these women take in making themselves financially dependent upon their husbands or partners, and Dutch business woman Marianne Zwagerman lashed out against women who “start a webshop with self-crocheted bibs, and think they have a career”. Such discussions also have their counterpart in academic debate, where (U.S.) mommy blogging, as well as womens’ crafting, have been scrutinized in communication studies, feminist studies, and beyond (e.g., Kido Lopez, 2009; Bratich & Brusch, 2011). In our paper, we relate these debates to the specific sphere of 91
92
ETMAAL 2013
BOOK OF ABSTRACTS
sewing mommy blogging, which has so far not been the topic of extensive academic study. By focusing on the Flemish-Dutch blog community, we also aim to extend the academic debate beyond the context of the United States, where mommy blogging has been discussed most vigorously. Our qualitative study of 25 popular sewing mommy blogs in the Netherlands and Flanders aims to provide insights into the socio-political dynamics and feminist politics of this cultural phenomenon. It does so based on a critical discourse analysis of these blogs, which explores what we consider to be the genre’s constitutive dimensions: (1) the genre’s intertextual relationship to earlier forms and norms of hegemonic motherhoods and femininities (including the notorious ‘yummy mummies’); (2) the aesthetic regime that can be seen to underline (both verbal and visual) communicative practices in this genre, and (3) the genre’s interdiscursive relationship with trends in other spheres of social life, most notably the movements of green politics and anti-consumerism, the debunking of romantic parenting in popular culture texts, and the upsurge in everyday and domestic forms of ‘micropolitics’. The results of the analysis of the blogs’ three dimensions point at a paradoxical ambivalence in the cultural practice of sewing mummy blogs. That is, the aesthetic and communicative practices on the blogs show an almost playful and highly reflexive mode of the bloggers simultaneously working within and rejecting the binaries of the public debate about sewing mums (“career-focused versus family-focused”, “feminist versus traditional”, etcetera). Yet, this ambivalence is indeed paradoxical. The blogs in fact enable the women to perform a particular mode of feminism / political activism, one in which highly complex ideological dilemmas (“challenging capitalism through consumption”, “celebrating quirkiness and idiosyncrasy by working within formats and templates”, “challenging sexism whilst reinforcing gendered patterns in design and clothing”) can be quite happily reconciled. We conclude by considering the feminist ramifications of this claimed form of micro-political engagement or activism. While the blogs, on the one hand, constitute a prime example of the circular problematic of feminism (see McRobbie, 2004) as well as an outlet for creative labour that is not acknowledged or credited in the global and capitalist economy, we argue that the sewing mummy blogs highlight the continuous inventiveness of the neoliberal regime of production and consumption.
Session #5 Friday 8 February 10:15 – 11:15 hrs. Chair: Katrien van Cleemput Location: Praag (M3-05) English session
MOBILE FITNESS APPS FOR PROMOTING PHYSICAL ACTIVITY ON TWITTER: THE #RUNKEEPER CASE - JEROEN STRAGIER AND PETER MECHANT Regular physical activity can reduce the probability of health problems such as obesity, type II diabetes and depression. That is why adults aged 18-65 years are advised a 30-minutes workout of moderate intensity 5 days a week. Unfortunately, many don’t meet these minimum requirements. In this paper we will look at the potential of Mobile Fitness Apps (MFAs) to promote physical activity 92
93
ETMAAL 2013
BOOK OF ABSTRACTS
and create exercise adherence. Afterall, although computer use may be related to a more sedentary lifestyle, online social networking and MFAs may just provide a way of promoting physical activity as many smartphone applications enable people to track fitness workouts and share these with online peers. “Just completed a … mi run/walk/ride…” is a frequently posted tweet on Twitter. Mobile fitness apps (MFA) have recently seen a steady increase in users and they afford a new, connected way of sharing the results of workouts. This paper focuses on why people want to share their workouts with their online social networks and applies this question to Runkeeper, a widely used MFA, and the hashtag #Runkeeper on Twitter. The results indicate that finding fellow runners, getting feedback and sharing activity information are the main reasons for sharing workouts on Twitter.
A BODY IMAGE APP: DEVELOPMENT AND EVALUATION OF A MOBILE APP FOR BODY IMAGE IMPROVEMENT AMONG ADOLESCENT GIRLS - JOLANDA VELDHUIS, ELLY A. KONIJN AND MARLOES SPEKMAN Media are blamed for negatively impacting adolescent girls’ body image (e.g., Grabe et al., 2008), while the very same media also are popular venues for providing information and positive experiences (Spettigue&Henderson, 2004). Thus, why not use media for improving one’s body image? Previous research shows that body-relevant information accompanying ideal-embodied models in magazine-settings may (re)direct body perceptions in women (Slater et al., 2012). Yet, thus far, hardly anything is known about applying such innovations in real-world settings or developing such messages within contexts that particularly address adolescent girls. A mobile device application might be an appropriate and ecologically valid setting for effectively redirecting negative effects of ideal-body exposure in adolescent girls. Hence, in the present study, a ‘body-image-app’ was developed and evaluated with adolescent girls. In line with entertainment education (Singhal&Rogers, 2002) and developmental insights on serious games for adolescents (Konijn&Nije Bijvank, 2009), previous research has shown positive results of games and apps for health improvement (Kato, 2010). Generally, interactive learning and playing serious games is intrinsically motivating and raises self-esteem (Gee, 2007; Griffiths, 2002). Especially for adolescents, including incentives in games and apps is important to keep them playing (Van Leijenhorst et al., 2010). Since rewards positively impact involvement and behavioral determinants (Zichermann&Cunningham, 2011), we included rewards in our app design to positively influence participants’ self-esteem. In regard to the app content, information to counteract media-induced negative body perceptions should accurately confirm media models’ weight status by attaching short, textual, information labels (Cowburn&Stockley, 2005) to the models’ pictures, which reduced body dissatisfaction in adolescent girls (authors). Thus, including appropriate weight information on media models (versus body-irrelevant information) as app content, is expected to raise body satisfaction. In sum, the study aimed to assess the appreciation and implementation of a body-image-app for adolescent girls (RQ1). Furthermore, we hypothesized (H1) that playing a body-image-app that incents adolescent girls for giving correct answers, increases their self-esteem. Finally, we expected (H2) that playing an informative body-image-app (i.e., informing about the weight status of media 93
94
ETMAAL 2013
BOOK OF ABSTRACTS
models) increases body satisfaction compared to playing a neutral body-image-app (i.e., bodyirrelevant quiz questions). Method. In a pretest-posttest experimental design, participants (N=210 adolescent girls; Mage=13.86, SDage=1.35) were randomly assigned to one of two experimental conditions of the body-image-app (body-relevant vs. neutral information). Designing the app was based on girls’ preferences in (serious) gaming and on Kaplan’s (2012) reference points for successfully developing effective interventions using new media. Both app versions displayed four ideal-embodied bikini models accompanied by quiz-like questions. In the body-relevant-information condition, the participants rated each models’ weight status (i.e., based on previous research). In the neutral condition, the models were accompanied by body-irrelevant questions about a famous singers-duo. Compliments followed correct answers, while incorrect answers allowed participants to try again. First, the impact of the apps on adolescent girls’ body (dis)satisfaction and self-esteem was tested (T0 and T1). Second, appreciation of the apps and various evaluation factors were assessed through an AppQuestionnaire. Based on innovation determinants of Fleuren et al. (2004) and the I-PEFiC model (Van Vugt et al., 2007), the AppQue used categorical items and open-ended questions, and tested three domains: innovation (i.e., information value, valence, and aesthetics), user (i.e., use and diffusion intentions), and use context (i.e., financial resources, time availability, and use environment). Results and conclusions. As expected (H1), paired-samples t-tests revealed that playing any of the body-image-app versions significantly improved self-esteem. However, we found no effect of correctly informing participants about the bikini models’ body shape on body (dis)satisfaction (H2), which contrasts previous findings on labeling models’ weight status. The results imply that providing positive incentives (even when these are body-irrelevant) accompanying thin-ideal exposure might yield protection for negative effects in increasing adolescent girls’ self-esteem. Results regarding RQ1 highlight several relevant insights. For example, regarding the innovation domain, the information app was considered significantly more informative, whereas the neutral app was perceived as significantly more clear, positive, and pleasant to play. Both app versions were equally involving, attractive, and creating a good feeling. Furthermore, the girls’ top-3 topics for apps are: fashion, looks, and feeling good - quite essential given our topic of interest. Results in the user domain ask for an extended app version. Results regarding the use context indicated apps as appropriate channels for targeting girls given their time and financial resources spend on apps. Importantly, answers to the open-ended questions also revealed pitfalls of communicating about body image by means of an app, such as boomerang effects. A complete overview and implications are presented at the Etmaal-conference.
WHEN COMPUTER-MEDIATED SUPPORT GROUP PARTICIPATION BACKFIRES: EFFECTS OF PESSIMISTIC SOCIAL COMPARISONS AND LACK OF PEER SUPPORT - ANIKA BATENBURG AND ENNY DAS Due to mixed findings on the effectiveness of computer-mediated support groups (e.g., Klemm et al., 2003; Rodgers & Chen, 2005; Fogel, 2004; Winefield et al., 2003), there is a need for research 94
95
ETMAAL 2013
BOOK OF ABSTRACTS
explaining why and when online health communities are beneficial for patients. Patients visit health communities for information, support and recognition in stories from others, which, in turn, empowers patients (Uden-Kraan, Drossaert, Taal, Seydel, & Laar, 2009; Barak, Boniel-Nissim, & Suler, 2008). However, patients also mention downsides, such as confrontation with threatening stories (e.g., Uden-Kraan et al., 2009). The present study aims to examine the effectiveness of online support group participation as a function of information seeking, receiving support, and social comparison. Receiving social support is an important determinant of patient wellbeing. Cancer patients seem to benefit from peer support, because they understand each other’s situation (Ussher, Kirsten, Butow, & Sandoval, 2006). We expect that active forum use may be particularly beneficial when it generates high levels of support. Patients who participate actively and receive higher levels of support from online peers feel better than active patients who perceive the support as insufficient. Patients also go online to find recognition in stories from peers (Uden-Kraan et al., 2008). Depending on whether patients compare themselves to others that are doing worse or better, patients may feel optimistic, frustrated, lucky, or threatened (Dibb & Yardley, 2006). Based on the social comparison literature, it is expected that active forum use may therefore be particularly beneficial for patients who use positive -but not negative- social comparisons. Finally, some patients have a higher need for information than others (Miller, 1987). High-monitors experience more distress than low-monitors because of their tendency to monitor and amplify threatening information (e.g., Miller, 1987; 1999). Therefore, it is expected that active forum participation backfires for high-monitors. Method A sample of 115 patients visiting an online breast cancer forum filled out an online questionnaire. Online forum participation was assessed by frequency of visits, average length of visits, forum contribution, and frequency of posts in the last four weeks. We assessed perceived emotional support and instrumental support (Brief COPE; Carver, 1997), social comparison (downward identification, upward contrast, upward identification, and downward contrast; Buunk et al., 1990) and monitoring (TMSI; Ong et al., 1999). Four scales measured patients’ wellbeing: depression (CESD10; Irwin, Artin & Oxman, 1999), breast cancer related concerns (Profile of Concerns about Breast cancer; Spencer et al., 1999), emotional wellbeing (FACIT-B; Brady et al., 1997), and anxiety (Dutch Mini-MAC; Watson et al., 1994). In addition, we measured participants’ working status, the amount of medical appointments the last 3 months, offline social support (FACIT-B; Brady et al., 1997), illness stage (Oldenburg, 2009), if patients were under treatment at the moment, and if patients received any psychological help. All scales were reliable (α > .70). Results Regression analysis was used to examine relationships between forum use, moderating factors (social support, social comparison, and monitoring), and measures of wellbeing. Active forum participation was positively correlated with upward identification, instrumental support, emotional support, monitoring, and the number of medical appointments. Overall, forum participation decreased emotional wellbeing, increased breast cancer related concerns, feelings of depression, and anxiety for patients who experienced lower levels of social support. Forum participation had no effects for patients who experienced high levels of support. Active forum participation also increased breast cancer related concerns for patients identifying with stories from peers doing worse, and had 95
96
ETMAAL 2013
BOOK OF ABSTRACTS
no effects for patients who did not. No interaction effect of monitoring and forum participation on wellbeing was found. Discussion Patients participated in online support groups because they are looking for positive stories from peers, emotional and instrumental support, and information. Despite the empowering processes provided by online forums, forum participation was in itself not associated with better psychological wellbeing, and sometimes even showed negative effects. Specifically, when patients perceived social support as insufficient, active participation had a negative impact on various measures of wellbeing. The perception of receiving insufficient support while participating quite actively may be disappointing. Furthermore, when patients identified with pessimistic stories, being more active in the online community also decreased wellbeing. Overall, these findings point to the importance of (dis)empowering processes in the online sphere: when patients feel not supported by their online peers or when they identify with the ‘wrong’ patient, forum participation backfires. A limitation of the present study is that it is cross-sectional. Future research is needed to confirm causality.
DATING A ROBOT IS LIKE DATING A MAN - A NEW PERSPECTIVE ON THE TURING TEST - MATTHIJS PONTIER, JOHAN HOORN, ELLY KONIJN AND RENS VAN DE SCHOOT Increasingly, humans communicate through technological devices, such as mobile phones and tablet PC’s. Insights from interpersonal communication are important for developing communication technologies. Certain communication technologies use these insights to mimic elements of human interpersonal communication, or even completely replace the human partner in conversational exchanges. With these developments, computer-mediated communication (CMC) is moving toward human-computer interaction (HCI). At the same time, with increasing knowledge about interpersonal communication and technological advances, HCI is becoming more social and moving toward CMC. Individuals respond socially toward computers. Humans, even experienced computer users, are inclined to treat their computers as largely natural and social and they tend to interact with their computers in affective ways. Nowadays, embodied computers are designed as socially interactive robots. They are not only technological supplementary tools for labor-intensive or hazardous tasks anymore, but gradually become part of our daily lives, fulfilling roles as, for example, pets, nurses, office assistants, tour guides, teachers, and even emotional companions (e.g., Sony’s AIBO robot dog, or MIT’s Kismet). A salient aspect of humanness in robots is the expression of emotions. However, compared to human affective complexity, affective communication of contemporary software agents and robots is still quite simple. Already in the Turing-Test, computer-mediated interpersonal communication was compared to HCI by means of “interrogators,” people who had to decide whether they were interacting with a machine or not. To our knowledge, Turing tests have not been used in computer-mediated conversations to evaluate whether individuals can discern affective communication coming from a machine or a human. For our present study, we implemented theories of human emotional intelligence to make a virtual human simulate emotional intelligence while communicating with its
96
97
ETMAAL 2013
BOOK OF ABSTRACTS
human users in a Turing test. The original Turing-Test is text-based but we enriched the communication by applying a virtual human in graphical format, called Tom, capable of speech and facial expressions. Because in mediated interpersonal communication emotions play a salient role, Tom’s communications were affect-laden. Our Turing test took the form of a speed-dating session in which we tested the performance of our Silicon Coppélia software. Female participants were randomly assigned to the virtual human Tom, who was either controlled by our software (HCI) or by a human confederate (CMC). In either case, participants were told that they interacted with a robot to see whether human performance could outdo that idea. Tom could give verbal as well as nonverbal cues. Results obtained with Bayesian structural equation modeling revealed that, even in an enriched and more sophisticated version of the classic Turing-Test, participants did not detect differences between the two versions of Tom, which is a nice tribute to the Turing Centenary of this year. Participants did not recognize significant differences in the cognitive-affective structure underlying the emotional behavior produced by our Silicon Coppélia software and that produced by human confederates. It must be noted, however, that in this experiment, interaction possibilities were limited (i.e., selected choice questions and answers) and richer expression forms should be considered in the future. We can conclude, as far as the testing capabilities of this experiment go, that we created a humanoid simulation of affect so natural that young women could not discern dating a robot from dating a man. These findings are relevant for communication scholars trying to understand human affective communication as well as for engineers who want to use our models for application development. Our computational model of affect can predict the mood and the possible affective reactions of one’s interaction partner. This may be especially useful as an Augmentative and Alternative Communication tool (AAC) for people with, for instance, autism, who typically have trouble estimating the emotional state of others. Additionally, our computational model can deepen the user experience of (serious) digital games, virtual stories, tutor and advice systems, telemedicine applications, such as coach or therapist systems, and social support groups, such as virtual patient communities. Furthermore, in settings where humans have to interact intensively with automated systems, the system can give sensor-data as input to our software to maintain a model of the emotional state of the user. This can enable the system to adapt the type of interaction to the user‘s needs.
97
98
ETMAAL 2013
BOOK OF ABSTRACTS
Session #6 Friday 8 February 11:45 – 12:45 hrs. Chair: Bas van den Putte Location: Praag (M3-05) English session
INFLUENCING THE CONVERSATION: EMOTIONS, CONVERSATIONAL VALENCE, AND BINGE DRINKING - HANNEKE HENDRIKS, BAS VAN DEN PUTTE AND GERT-JAN DE BRUIJN Health campaign effects may be improved by taking interpersonal communication processes into account. That is, many studies have demonstrated that conversational occurrence (whether people talk about a specific health topic) is important for the effects of health campaigns (e.g., Geary et al, 2007). Moreover, several recent studies have shown that conversational valence (i.e., whether people converse positively or negatively about the subject at hand) is potentially even more relevant than mere conversational frequency for the effects of health messages (e.g., Hendriks, De Bruijn, & Van den Putte, 2012). Given the importance of conversational valence, research into predictors of a desired conversational valence is very valuable. There is some indication that emotions influence the valence of conversations (e.g., Dunlop, Kashima, & Wakefield, 2010); however, more research is needed to ascertain this claim. Further, which specific emotions are especially relevant for conversational valence is not yet known. In the current study, by employing an experimental, pretest-posttest, randomized exposure design, we specifically designed and tested different types of emotional messages (fear, disgust, and humor appeals) in order to discern whether, and if so which, emotions are especially effective at inducing a desired conversational valence and subsequent healthy behaviors. We focus on alcohol abuse and binge drinking given the fact that this is an extensive societal problem resulting in many negative consequences (e.g., Geels et al., 2012). Based on literature, we expect that: (H1) Exposure to an emotional appeal more strongly elicits congruent emotions than alternative emotions. For instance, we expect that our fear ad more strongly induces fear emotions than disgust or humor related emotions. Further, we hypothesize: (H2a) an increase in the emotion fear is related to a more negative conversational valence about alcohol, (H2b) an increase in the emotion disgust is related to a more negative conversational valence about alcohol, (H2c) an increase in the emotion humor is related to a more positive conversational valence about alcohol, and (H2d) the emotion fear more strongly induces a negative conversational valence than the emotion disgust. Lastly, we expect that (H3) a more negative conversational valence about alcohol elicits a more negative attitude towards binge drinking, induces a lower intention to binge drink and decreases actual binge drinking behavior. At baseline, 208 participants provided information concerning their binge drinking attitudes, intentions, and behaviors. Two weeks later, participants were randomly shown one of four antialcohol messages (a fear appeal, a disgust appeal, a humor appeal, or an informational appeal) which were specifically designed and pre-tested for the purpose of this study. After this, participants discussed the topic alcohol and binge drinking in dyads. Next, emotions, conversational valence, and binge attitudes, intentions, and behaviors were measured. 98
99
ETMAAL 2013
BOOK OF ABSTRACTS
The three main findings were as follows. First, an increase in the emotion fear induced a more negative conversational valence about alcohol. The emotions disgust and humor were unrelated to conversational valence. Second, fear was most strongly induced by exposure to a disgusting antialcohol message and was lowest after watching a humorous appeal. Third, a more negative conversational valence elicited healthier binge drinking attitudes, intentions, and behaviors. Hereby, this study sheds light unto several important issues. Firstly, it seems that whether people talk about health issues can be predicted by multiple emotions (e.g., fear, disgust, and humor; Peters et al., 2009); however, the valence in which people talk is influenced by fewer emotions. In particular, our study showed that solely the emotion fear was linked to conversational valence. Potentially the arousing nature of the emotion fear can account for these differences (Berger & Milkman, 2011). Secondly, in line with research linking disgust and fear (Leshner, Vultee, Bolls, & Moore, 2007), our findings indicate that a disgusting ad is an especially effective trigger of fear and subsequent desired conversational valence. In contrast, exposure to a humorous appeal decreased feelings of fear. Thus, in future health promotion attempts it seems worthwhile to include disgusting elements, and to avoid using humorous elements, in order to induce a desired conversational valence. Thirdly, our findings show that conversational valence affects a variety of health persuasion outcomes, ranging from health attitudes to actual health behaviors. Concluding, this study suggests that health researchers should take interpersonal processes into account when studying health campaign effects. Furthermore, health campaign planners should pay special attention to the emotional characteristics of their ads and should try to induce a desired conversational valence in order to increase healthy conduct.
INTERVENING OR INTERFERING? THE INFLUENCE OF POSITIVE AND NEGATIVE INJUNCTIVE AND DESCRIPTIVE NORMS MESSAGES PROTECTIVE BEHAVIORS IN ALCOHOL CONSUMPTION CONTEXTS -
SAAR MOLLEN, RAJIV N. RIMAL, ROBERT A. C. RUITER, SU AHN JANG AND GERJO KOK Excessive alcohol consumption among college students is a serious issue, as it has been linked to multiple negative outcomes. One of the reasons underlying excessive drinking among college students is a discrepancy between perceived and actual drinking norms. Students share the tendency to overestimate how much others drink (i.e., descriptive norm), as well as the overall level of approval of heavy drinking (i.e., injunctive norm). The “social norms approach,” an intervention method designed to reduce heavy drinking, focuses on disseminating positive norms by communicating that most students either do not drink, or drink responsibly. Or that the uniformity of approval of drinking is not as high as one might think. The communication of lower drinking norms is thought to offset overestimations of drinking behavior among students and create new norms that will in turn guide alcohol consumption. This is a very popular approach, however, these interventions have shown mixed success. While some have proven to be successful, others have failed to encourage significant changes in alcohol consumption. Turner, Perkins and Bauerle (2008) conducted an intervention that did produce significant changes in alcohol consumption and negative consequences related to alcohol consumption. In addition to disseminating messages that were designed to reduce misperceptions of drinking prevalence and perceived approval, they communicated messages about protective behaviors, such as asking friends 99
100
ETMAAL 2013
BOOK OF ABSTRACTS
who had been drinking excessively to slow down, or not leaving a friend who had been drinking alone, and that these interventions among friends were in fact the norm. These findings illustrate that a promising approach to reduce negative consequences related to alcohol consumption might be to communicate the normative nature of intervening actions when partying with friends. Messages conveying the high prevalence of intervening behaviors in alcohol consumption contexts might help people to overcome the idea that one is interfering and could increase motivation to actually engage in intervening behaviors. In the current study it was therefore investigated whether positive and negative descriptive and injunctive norms with regard to protective behaviors in alcohol consumption contexts, would affect people’s motivation to engage in intervening (e.g., ask a friend to stop drinking) and non-intervening behaviors (e.g., let friends make their own drinking decisions). A total of 240 university students participated in an online experiment with a post-test. Participants were presented with a positive or negative descriptive (i.e., what most others do) or injunctive (i.e., what most others approve or disapprove of) norm message. The positive norm messages described the normative nature of intervening behaviors (i.e., monitoring a friend’s drinking behavior and making sure friends get home safely), while the negative norm messages described the normative nature of non-intervening behavior (i.e., letting friends decide for themselves how much they want to drink and when to go home). After exposure to the norms message participants were asked to indicate their motivation to engage in intervening and non-intervening party behaviors. Four weeks after participating in the main experiment, participants were asked to indicate whether in the past four weeks they had engaged in any protective behaviors, such as bringing a friend home after a party or telling him or her to slow down their drinking. Results show that positive social norms resulted in a stronger motivation to engage in intervening behaviors, compared to negative social norms. Additionally, those who were presented with positive normative information were less prepared to engage in non-intervening behaviors, compared to those who read a negative norms message. Findings with regard to actual behavior at follow-up showed effects of injunctive, but not descriptive, norms on protective behaviors during the consecutive month. Those who read a negative injunctive norm message engaged in less protective actions, than those who read a positive injunctive norm. This is in line with prior findings regarding the transsituational influence of social norms, which shows that injunctive, but not descriptive, norms influence behavior over different situations. It has to be noted, however, that effects of these messages on (non-)intervening motivation and behavior were found for actions regarding others’ drinking, but not motivation and behavior regarding going home after a party. Findings from this study indicate that communicating that it is normal or that people approve of intervening in alcohol consumption contexts can bring them to the realization that they are intervening instead of interfering. The behavior that results from this realization can be an important tool in reducing negative consequences related to drinking and can guide future social norms interventions to reduce alcohol consumption among college students.
100
101
ETMAAL 2013
BOOK OF ABSTRACTS
SELF-AFFIRMATION BEFORE EXPOSURE TO PERSUASIVE HEALTH COMMUNICATIONS PROMOTES ANTICIPATED REGRET AND HEALTH BEHAVIOR CHANGE - GUIDO M. VAN KONINGSBRUGGEN, PETER R. HARRIS AND ANJES J. SMITS Introduction Anticipated regret, the prospect of feeling regret after performing or not performing a behavior, has emerged as a significant predictor of health intentions and behavior (e.g., Sandberg & Conner, 2008). Accordingly, it has been argued that persuasive health communications that elicit anticipated regret could be a viable strategy to promote health behavior change (e.g., Morison, Cozzolino, & Orbell, 2010). Unfortunately, health messages often provoke defensive reactions from target audiences (Dillard & Shen, 2005), and an explicit focus on inducing regret may unintentionally strengthen such resistance. From the perspective of self-affirmation theory (Steele, 1988), health messages that encourage people to change aspects of their lifestyle threaten their sense of self-integrity, and trigger a need to restore this sense. One way to restore this sense is to respond defensively to such messages. The theory further proposes, however, that affirming unrelated domains of self-worth (e.g., being a friendly person) should reduce the need to respond defensively to subsequent threats in another domain (e.g., health). Indeed, self-affirmation (reflecting upon cherished values) before message exposure reduces defensive reactions and promotes health behavior change (Harris & Epton, 2009). Moreover, self-affirmation triggers higher levels of construal (Wakslak & Trope, 2009), which makes people adopt a longer term perspective when processing information (Trope & Liberman, 2003). Intriguingly, people who focus more on the future implications of their behavior are more likely to experience feelings of anticipated regret (e.g., Morison et al., 2010). Taken together, these lines of research suggest that self-affirmation could boost anticipated regret in response to persuasive health communications, thereby promoting healthy intentions and behavior change. However, a direct test of this fascinating hypothesis is lacking. We therefore conducted two studies to provide such a test. Methods We randomly allocated female participants to a self-affirmation or control task before reading a message regarding the health-promoting effects of fruit and vegetables. The health message was adopted from Epton and Harris (2008). In the self-affirmation condition, participants completed items presumably measuring their personal strengths, which focused their minds on important selfvalues (e.g., value of humanity; “I am never too busy to help a friend”). In the control condition, participants completed the same set of items judging the qualities of a celebrity instead of their own strengths (see Napper, Harris, & Epton, 2009, for a validation of this manipulation). Anticipated regret (2 items) and intentions (3 items) were measured immediately after exposure to the health message. We also measured baseline fruit and vegetable consumption at the beginning of the studies. Study 1 (N = 42) was conducted in the laboratory and designed to provide an initial test whether selfaffirmation would boost anticipated regret and intentions to behave healthily. We also examined effects on behavior using a computerized task in which participants had to repeatedly choose
101
102
ETMAAL 2013
BOOK OF ABSTRACTS
between fruit or vegetables and high-calorie snacks. Study 2 (N = 91) was designed to provide a second test of our hypothesis outside the laboratory and comprised two points of measurement: a treatment session and a one-week follow-up in which participants reported their fruit and vegetable consumption in the last 7 days. In both studies we tested if any changes in intentions and behavior induced by self-affirmation were mediated by differences in anticipated regret (and/or intentions) between conditions following exposure to the health message. Results Regression analyses tested the effects of condition (control vs. self-affirmed) on the dependent measures. To control for individual differences in fruit and vegetable consumption prior to the manipulation, baseline behavior was entered as a control variable in the analyses. Self-affirmation increased anticipated regret and intentions to consume fruit and vegetables, and the effects of self-affirmation on intentions were mediated by anticipated regret (Studies 1 and 2). While there was no effect on the food-choice task (Study 1), self-affirmation increased consumption during one week following the intervention in Study 2. This effect was mediated by intentions. Discussion Self-affirming people before exposing them to a persuasive health communication boosts anticipated regret and promotes health behavior change. Previous research identified anticipated regret as an important predictor of health intentions and behavior. While explicitly triggering anticipated regret using persuasive health communications might not always be appropriate, self-affirmation thus appears to provide an indirect strategy to profit from the benefits of eliciting this construct in target audiences. Moreover, by demonstrating the mediating role of anticipated regret, this research provides novel insights into how self-affirmation promotes the formation of healthy intentions that lead to improvements in health behavior.
FLOSSING, IMPLICIT ATTITUDES AND MESSAGE FRAMING EFFECTIVENESS - GERT-JAN DE BRUIJN, BAS VAN DEN PUTTE AND JONATHAN VAN 'T RIET Theoretical Background Message framing is considered to be an important persuasive communication strategy to promote healthy behaviours and to discourage unhealthy behaviours. Recent meta analyses, however, have cast some doubt on the applicability of two types of messages (i.e. gain frames and loss frames) to change two types of behaviour (respectively, prevention and detection behaviours). Subsequently, moderator variables are suggested to be taken into account when designing and testing the effectiveness of message framing interventions. Recent studies have provided evidence that various moderator variables impact on message framing effectiveness, such as self-efficacy, prior experience, risk perception and behavioural automaticity. No evidence, however, exists on how pre-existing implicit attitudes may impact the effect of framed message exposure. The present study thus investigated how implicit attitudes towards the target behaviour may moderate the effectiveness of framed messages. We focused on dental flossing behaviour. Methods 102
103
ETMAAL 2013
BOOK OF ABSTRACTS
Three wave data were collected in 142 undergraduate students who participated either for course credits or 10 euros. Baseline data included implicit attitudes towards flossing behaviour (assessed using Millisecond software) and flossing behaviour and intentions. One week later, participants were randomly exposed to either a gain framed message or a loss framed message and again reported flossing behaviour and intentions. Finally, one week after message exposure, participants reported flossing behaviour and intentions again. Data were analysed using repeated measures analysis for a 2 (message frame) x 3 (baseline implicit attitude: negative, mean, positive) design. Results For intentions, there was a significant of time, F(2,149)=4.73, p=.01, but nonsignificant effects of time*condition, F(2,149)=1.39, p=.26, of time*implicit attitude, F(2,149)=0.57, p=.01, and of time*implicit attitude*condition. Thus, the message intervention did not affect intentions towards flossing behaviour. For flossing behaviour, there were nonsignificant effects of time*condition, F(2,149)=0.28, p=.76, of time*implicit attitude, F(2,149)=0.28, p=.89, but there was a significant time*condition*implicit attitude interaction, F(2,149)=4.73, p=.001. Paired comparisons showed that the loss-framed message was most effective for behavioural change when participants already had positive implicit attitude towards flossing behaviour at baseline. Discussion The present study indicates that prevention behaviours such as flossing are not changed by gainframed messages, as theory would predict. In fact, loss framed messages were more effective in changing flossing behaviours. However, this effect was only observed amongst those who already held favourable implicit attitudes towards flossing behaviour. It is concluded that message framing theory may need to be broadened beyond the typical 2 (type of behaviour: detection vs prevention) x 2 (type of message: loss vs. gain) dichotomy by including additional variables that impact framed message effectiveness.
Session #7 Friday 8 February 11:45 – 12:45 hrs. Chair: Reint Jan Renes Location: Auckland (M3-04) English/Dutch session
CYBERBULLYING ON SOCIAL NETWORK SITES: AN EXPERIMENTAL STUDY INTO BYSTANDERS’ BEHAVIORAL INTENTIONS TO HELP THE VICTIM OR REINFORCE THE BULLY - SARA BASTIAENSENS, ILSE DE BOURDEAUDHUIJ, ANN DESMET, KAROLIEN POELS, KATRIEN VAN CLEEMPUT AND HEIDI VANDEBOSCH Cyberbullying is a contemporary problem amongst European adolescents (Livingstone, Haddon, Görzig, & Olafsson, 2011). As in traditional bullying, cyberbullying can be enacted directly (e.g. sending an insulting message to a person), or indirectly via one or more third parties (e.g. creating a hate website about a person) (Dooley, Pyżalski, & Cross, 2009). In both instances, (potential) 103
104
ETMAAL 2013
BOOK OF ABSTRACTS
bystanders can play an important role in the (cyber)bullying process. They have the ability to reinforce the bullying, but they can also take sides with the victim in an attempt to stop the bullying or prevent negative effects (Salmivalli, 2010). With regard to the determinants of bystanders’ behavior in cases of cyberbullying, several research gaps remain. To date, cyberbullying research has mainly focused on bystanders’ cognitive mechanisms to rationalize and justify harmful acts, including moral disengagement (Pornari & Wood, 2010) and ‘blaming the victim’ (Varjas, Talley, Meyers, Parris, & Cutts, 2010). Little is known about the influence of the behavior of other bystanders. Especially in cyberbullying through social network sites (SNS), the presence of many bystanders is probable and the actions of others (assisting the bully or defending the victim) could impact bystander behavior. Additionally, the identity of other bystanders needs further investigation. Experimental research on bystanders of a videotaped violent attack revealed that other bystanders are only influential if they are regarded as group members (Levine, Cassidy, Brazier, and Reicher, 2002). Therefore, it would be interesting to look at differences in behavioral intentions when these acting bystanders are mere acquaintances versus good friends. Furthermore, the perceived severity of the cyberbullying incident appears to have an influence on bystander reporting (DeSmet et al., 2012) and should therefore be taken into account. In order to address the aforementioned issues, we set up an experiment to study the behavior of bystanders of cyberbullying on SNS. We use insights from the Bystander Effect by Latané and Darley (1970) and from peer pressure experienced by bystanders (Levine et al., 2002; Nickerson, Mele, & Princiotta, 2008) to develop hypotheses on the influence of group dynamics on bystanders’ intentions to perform helping behavior (or not) towards the victim. Approximately 500 Flemish students (12-13 years old) with different educational backgrounds will participate. The experiment involves a scenario that reports a cyberbullying incident on a SNS, involving one perpetrator and one victim, and bystanders reacting upon it. The participant has to imagine that he or she encounters this incident as a bystander. In the scenario, the severity of the cyberbullying incident (low versus high), the visible behavior of the other bystanders (assisting the bully versus defending the victim), and the identity of these bystanders (acquaintances versus good friends) are manipulated in a betweensubjects design. Afterwards, the respondents’ behavioral intentions are measured, including a whole range of behaviors to assist the bully or help the victim, online or offline and public (visible to others) or private (one-to-one). Covariates include involvement with the situation, empathic tendencies (de Corte et al., 2007), moral disengagement (Hymel, Rocke-Henderson, & Bonanno, 2005) and real life involvement in (cyber)bullying. We are currently performing pretests to fine-tune the different factors manipulated in the scenario. The experiment is scheduled for November - December 2012, which means we will be able to report preliminary results in a conference paper, followed by an update in a presentation at the conference. This experiment can give valuable insights into the role and behavior of bystanders of cyberbullying on SNS. Results of this study can further guide the creation of interventions against cyberbullying. Firstly, we investigate the influence of group-processes (e.g. peer pressure) on bystanders of cyberbullying on SNS. If these group-processes are confirmed, an intervention could use this knowledge to teach youth how to resist (negative) peer pressure, in order to avoid assisting the bully and increase helping behavior towards the victim. Secondly, if severity perception would lead to more helping behavior, this could be included in an intervention by trying to make the overall perception of cyberbullying more severe. The insights from this experiment might also guide new 104
105
ETMAAL 2013
BOOK OF ABSTRACTS
research: future studies can, for instance, test whether intended “helping” actions of bystanders, are really effective (e.g. does a victim experience less distress if he or she receives a supporting e-mail?)
MODELLING PROXIMAL DETERMINANTS OF CYBERBULLYING PERPETRATION - SARA PABIAN AND HEIDI VANDEBOSCH This paper starts from the observation that research on cyberbullying perpetration has paid relatively few attention to proximal determinants of this behavior. It therefore tests the value of a model that departs from variables representing the beliefs underlying the central concepts of the Theory of Planned Behavior (TPB) to explain cyberbullying intention and behavior. The data stem from a longitudinal study amongst 1,606 students (11-17 years). A SEM analysis reveals that these beliefs account for respectively 88.1%, 41,6% and 31.5% of the variance in subjective norm (SN), perceived behavioral control (PBC) and attitude (A). The direct measurements of SN, A and PBC account for 35.1% of the variance of the intention to cyberbully and intention, in turn, explains 23% of the reported cyberbullying behavior six months later. We conclude that the extended TPB model provides a detailed insight into the proximal determinants of cyberbullying, which benefits future cyberbullying interventions.
THE EFFECT OF PLAYING ADVERGAMES THAT PROMOTE ENERGY-DENSE SNACKS OR FRUIT ON ACTUAL FOOD INTAKE AMONG CHILDREN - FRANS FOLKVORD, DOESCHKA ANSCHUTZ, MONIEK BUIJZEN AND PATTI VALKENBURG Previous studies have focused on the effects of television advertising on the caloric intake of children. However, the rapidly changing food marketing landscape requires research to measure the effects of nontraditional forms of marketing on the health-related behaviors of children. The main aim of this study is to examine the effect of advergames that promote energy-dense snacks or fruit on children’s ad libitum snack and fruit consumption and to examine whether this consumption differs according to brand and product type (energy-dense snacks and fruit). The second aim is to examine whether advergames can stimulate fruit intake. We used a randomized between-subject design with 270 children (age: 8–10 y) who played an advergame that promoted energy-dense snacks (n = 69), fruit (n = 67), or nonfood products (n= 65), or were in the control condition (n = 69). Subsequently, we measured the free intake of energy-dense snacks and fruit. The children then completed questionnaire measures, and we weighed and measured them. The main finding is that playing an advergame containing food cues increases general caloric intake, regardless of the advertised brand or product type (energy-dense snacks or fruit), and this activity particularly increases the intake of energy-dense snack foods. Children who played the fruit version of the advergame did not eat significantly more fruit than those in the other groups. The findings suggest that playing advergames that promote food, including either high energy-dense snacks or fruit, increases the caloric intake of children.
105
106
ETMAAL 2013
BOOK OF ABSTRACTS
SNOEPEN VOOR DE TELEVISIE: ZELFBEHEERSING NEUTRALISEERT EFFECT TELEVISIERECLAME HENK WESTERIK, LUC KIPMULDER, HANS LEMMENS, RUBEN KONIG, ESTHER ROZENDAAL AND MONIEK BUIJZEN Achtergrond en theorie. In toenemende mate is onder kinderen, adolescenten en volwassenen wereldwijd sprake van overgewicht. Uit eerder onderzoek komt naar voren dat blootstelling aan televisie en televisiereclame hiermee positief samenhangt. Bij het ontstaan van overgewicht spelen problemen op het gebied van zelfbeheersing (self-control) een mogelijke rol en in deze studie wordt gekeken in hoeverre televisiereclame hierop van invloed is. Zelfbeheersing wordt in deze studie opgevat als het vermogen een uitgestelde beloning te verkiezen boven een onmiddelijke beloning en onderzocht wordt of dit de effecten van reclame modereert. Methoden. Op 9 Nederlandse scholen werd een experiment uitgevoerd met in totaal 120 kinderen (leeftijd: 8-9). Proefpersonen werden at random toegewezen aan een conditie waarin zij keken naar een videofragment met reclames (inclusief enkele snoepreclames) of aan hetzelfde videofragment zonder reclames. Tegelijk werden ze toegewezen aan de conditie waarin een beloning voor niet snoepen tijdens het televisiekijken werd beloofd, of juist niet. Onderzoeksresultaten werden geanalyseerd met ANOVA, met directe snoepconsumptie (gemeten als het aantal gegeten M&M’s tijdens het kijken naar een 15 minuten durend videofragment) als afhankelijke variabele en ‘Motivatie tot zelfbeheersing’ en ‘Blootstelling aan reclames’ als verklarende factoren. Resultaten. Uit het onderzoek blijkt dat het in het vooruitzicht stellen van een beloning bijdraagt de eventuele effecten van een reclameblok op het gaan snoepen neutraliseert. Reclame heeft alleen effect op kinderen die niet gemotiveerd worden tot zelfbeheersing. Daarnaast heeft beloning op zichzelf ook een beschermend effect. Discussie. Uit het onderzoek blijkt dat een programmaonderbreking dat het motiveren van kinderen tot zelfbeheersing helpt om de invloed van mediagebruik te weerstaan. Voorlopig blijft echter nog onduidelijk of dit beschermde effect optreedt bij de blootstelling naar reclames in het algemeen of alleen/ook/juist niet bij reclames waarin snoepgoed wordt aangeprezen.
106
107
ETMAAL 2013
BOOK OF ABSTRACTS
JOURNALISM Session #1 Thursday 7 February 14:05 – 15:05 hrs. Chair: Karin Raeymaeckers Location: Shanghai (M2-12) Dutch session
‘FACEBOOK, DAAR MOETEN WE IETS MEE’ INTERACTIE TUSSEN NIEUWSMEDIA EN HET PUBLIEK OP FACEBOOK - SANNE HILLE In het laatste kwartaal van 2012 telt Nederland ruim 7 miljoen gebruikers op Facebook, wat erop neerkomt dat bijna 50% van de Nederlandse internetters een Facebook account heeft. Om een groter en breder publiek te trekken heeft de meerderheid van de Nederlandse nieuwsmedia ook de weg naar Facebook weten te vinden. Nieuwsmedia gebruiken het sociale medium om content te publiceren, content te laten verspreiden, een nieuw en jonger publiek te bereiken, meer bezoekers naar de eigen site te trekken en de band met het publiek te versterken door met hen te interacteren. Facebook-gebruikers kunnen op berichten reageren die nieuwsmedia op hun Facebook pagina plaatsen. Uit eerder onderzoek (Hille & Bakker, 2012) blijkt dat nieuwsmedia verschillende strategieën gebruiken en daarmee verschillende doelen willen bereiken. Door berichten te liken, sharen en aanbevelen helpen Facebook-gebruikers nieuwsmedia echter voornamelijk met de distributie van de journalistieke content terwijl interactie tussen (journalisten van) nieuwsmedia en het publiek nog nauwelijks lijkt plaats te vinden. De vraag is daarom of het gebruik van Facebook door nieuwsmedia inhoudelijk ook iets oplevert. Daarom zal dit onderzoek duidelijkheid trachten te verschaffen over de manier waarop nieuwsmedia en het publiek op Facebook met elkaar interacteren en over de inhoudelijke bijdrage die het publiek levert aan de journalistiek middels deze interacties. Ander onderzoek naar het gebruik van Facebook door media en het publiek (Hermida et al., 2011; Van Dijck, 2012; Hille & Bakker, 2012) heeft nog geen inzicht gegeven in de mate en kwaliteit van de interactie tussen journalistiek en publiek. De inzet van Facebook om publiek te laten reageren op nieuwsmedia is een betrekkelijk nieuw gegeven en is er dan ook nog geen systematisch onderzoek gedaan naar de gevolgen hiervan. Nieuwsmedia laten het publiek echter al langere tijd op hun eigen platform reageren. Reageren op nieuws is de meest voorkomende manier om publiek te laten participeren in de journalistiek maar is in vergelijking met andere vormen van publieksbijdrages de minst gewaardeerde vorm (WahlJorgensen et al., 2009). Vooral anonieme en niet gemodereerde reacties worden als kwalitatief minderwaardig beschouwd en leiden tot irritatie bij zowel publiek als redacties (Van Sloten, 2011). Het reageren op Facebook of het reageren op de eigen website met behulp van een Facebook account, zoals dat bij de Leeuwarder Courant, Dagblad van het Noorden en EenVandaag al gebeurt, 107
108
ETMAAL 2013
BOOK OF ABSTRACTS
zouden dan ook voor meer inhoudelijke en afgewogen reacties zorgen (Pleijter, 2011). Voor dit onderzoek zijn kwalitatieve inhoudsanalyses uitgevoerd op een sample van tien nieuwsmedia (voortbouwend op voorgaand onderzoek) die actief op Facebook aanwezig zijn: NRC.next, Trouw, de Volkskrant, AD, Nu.nl, Hart van Nederland, RTV Rijnmond, VPRO Tegenlicht, EenVandaag en De Groene Amsterdammer. In de week van 17 tot 26 september zijn van deze tien nieuwsmedia alle Facebook- berichten over ‘ProjectX’ in Haren verzameld. Deze berichten en reacties zijn vervolgens geanalyseerd met behulp van de kwalitatieve analyse software van Maxqda en het stappenplan voor inhoudsanalyses van Hansen et al. (1998). Bij de analyse is gekeken of er wel of niet inhoudelijk wordt ingaan op het bericht, of mensen op elkaar reageren, tips aan de redactie geven, wordt gescholden, beledigd of vooral oneliners en meningen worden geventileerd. De eerste resultaten van het onderzoek maken duidelijk dat geen van de nieuwsmedia of redacteuren na het plaatsen van een bericht zich nog in de discussie mengen. Nieuwsmedia roepen het publiek wel op om op geplaatste berichten te reageren. Zij doen dit door in de inleiding van de nieuwsberichten het Facebook-publiek vragen te stellen. De reacties bij de berichtgeving over ProjectX in Haren lopen per medium sterk uiteen. Bij de Volkskrant en Nu.nl wordt vooral inhoudelijk op de berichtgeving ingegaan. Zij reageren daarbij ook weer op elkaar waardoor er een discussie ontstaat. De reacties op het AD bevatten daarentegen veel meer scheldwoorden en oneliners waarbij de ‘feestgangers’, politiek maar vooral de media het moeten ontgelden. Uit de analyse wordt duidelijk dat de functie van het kunnen reageren op de berichten van nieuwsmedia op Facebook vooral een uitlaatklep voor meningen en opvattingen is, zonder dat er inhoudelijk wordt bijgedragen in de vorm van nieuwe invalshoeken of scoops voor een journalistiek verhaal.
THE MEDIUM IS THE COMMENT? DE INVLOED VAN HET NIEUWSPLATFORM OP DE LEZERSCOMMENTAREN - FANNY MAENHOUT AND MICHAEL OPGENHAFFEN Publieksparticipatie wordt beschouwd als een van de belangrijkste kenmerken van de online omgeving. Ook binnen journalism studies in Vlaanderen, Nederland en daarbuiten is er de voorbije jaren heel wat aandacht uitgegaan naar de deze vorm van interactiviteit, en dan meer bepaald naar de manier waarop nieuwsconsumenten interageren met het nieuws, met de journalist, en met elkaar (e.g., Hermida & Thurman, 2008; Paulussen, 2011). Het is vandaag de dag als nieuwsconsument immers mogelijk om online publiekelijk te reageren op een nieuwsartikel of om de journalist via bijvoorbeeld e-mail persoonlijk te bereiken. Professionele journalisten zijn echter erg bezorgd over de kwaliteit van de nieuwscommentaren en de vaak haatdragende taal die wordt gebruikt (Domingo et al., 2008). Journalisten zien de online lezerscommentatoren als oppervlakkiger en agressiever dan de oudere vormen van lezersparticipatie zoals via lezersbrieven (Reich, 2011). Als gevolg hiervan hebben verschillende nieuwsmedia hun discussiefora of reactiemogelijkheden op de nieuwssite afgesloten of een systeem van verplichte registratie voorzien. Dit onderzoek gaat dieper in op welke manier lezersreacties beïnvloedt worden door de context van het nieuwsartikel. Meer specifiek onderzoeken we of het platform waarop de interactiviteit tussen lezer en journalist plaats vindt een invloed heeft op de inhoud van een reactie. Onderzoek heeft immers aangetoond dat mediaplatformen een specifieke medialogica hebben met bijhorende 108
109
ETMAAL 2013
BOOK OF ABSTRACTS
kenmerken en kwaliteiten die een invloed hebben op de informatieoverdracht, de feedback en de taal (Daft & Lengel, 1984). Een belangrijk kenmerk is de graad van anonimiteit. Zo kan een vermindering van de sociale aanwezigheid leiden tot groepspolarisatie wat dan weer leidt tot extremere meningen en gedrag. Wanneer anonimiteit misbruikt wordt, kan dit leiden tot extreem taalgebruik (Coffey & Woolworth, 2004). Precies omdat de graad van anonimiteit verschillend is op de beschikbare online platformen, onderzoeken we of het gebruikte platform een invloed heeft op de inhoud van de lezerscommentaren. Methode Om de verschillende platformen en de manier waarop deze platformen door de nieuwsconsumenten gebruikt worden om te reageren te vergelijken, analyseren we in deze studie de reacties op de artikelen van de ombudsman van De Standaard. Sinds 15 juni 2011 publiceert Tom Naegels als ombudsman elke woensdag zijn stukken in de papieren krant en op de blog-sectie van de digitale krant. Bovendien legt de ombudsman op zijn Facebookprofiel een link naar deze artikelen en is hij ook op Twitter actief. Tot slot is er ook de mogelijkheid om via e-mail te reageren. We hebben voor deze case study uit de teksten die tussen 1 januari 2012 en 30 juni 2012 verschenen zijn vijf ombudsartikelen geselecteerd waarvan we telkens de reacties via de weblog, e-mail, Facebook en Twitter verzameld en geanalyseerd hebben. We hebben hierbij rekening gehouden met mogelijke overlap van inhoud en alleen teksten met voldoende reacties opgenomen in ons steekproef, dit om een vergelijking te kunnen maken tussen de reacties onderling op basis van het platform. Voor de dataverzameling van e-mails hebben we de medewerking gekregen van Tom Naegels zelf die ons alle mailreacties per artikel bezorgd heeft. Van elke reactie analyseerden we de graad van anonimiteit, het frame, het doel van de reactie en het taalgebruik. Resultaten De resultaten van de lezersreacties via de verschillende platformen laten zien dat er in het algemeen gezien weinig anonieme reacties verstuurd werden. Wel werden er vooral bij de reacties op de weblog afgekorte namen gebruikt, wat voor een bepaalde graad van anonimiteit zorgt. Reacties die via e-mail verstuurd werden, waren het meest negatief waren, terwijl deze vorm van communicatie dichter bij face-to-facecommunicatie ligt dan bijvoorbeeld de weblog en dus minder anoniem is dan de andere vormen. Ook via de blog werden er heel wat negatieve reacties verstuurd en was het aandeel positieve reacties het kleinst van alle platformen. Op Facebook en Twitter zien we vooral positieve reacties. Lezers lijken zich daar bewust van het feit dat hun reacties openbaar zijn en dat hun netwerk kan meelezen. Wat functie betreft waren de meeste reacties vooral bedoeld als losse opmerking, al zagen we vooral bij de weblog veel reacties als onderdeel van een debat onder de lezers en de ombudsman. Op Twitter werden de reacties dan weer vooral gedeeld, zonder toevoeging van een eigen reactie. Wat taalgebruik betreft, zien we dat Facebook het meest positieve platform is. De weblog heeft het grootste aandeel negatieve taal. Tijdens de presentatie gaan we dieper in op de verschillen tussen de verschillende platformen, de mogelijke oorzaken hiervan en implicaties voor de (online) journalistiek.
109
110
ETMAAL 2013
BOOK OF ABSTRACTS
REGIONALE MEDIA, DEMOCRATIE EN BURGERSCHAP - NIEK HIETBRINK AND LIESBETH HERMANS Regionale media worden belangrijke taken toegedicht in de samenleving. Hierbij wordt o.a. gewezen op een bijdrage aan de maatschappelijke cohesie en participatie, een onafhankelijke informatievoorziening ten dienste van burgers en samenleving en een bijdrage aan lokale en provinciale democratie. Kortom, media worden geacht een bijdrage te leveren aan burgerschap en democratie. Wanneer men spreekt over rol die media in een democratie spelen denkt men in eerste instantie aan de politieke berichtgeving en de berichtgeving over verkiezingen. Burgerschap komt echter ook buiten verkiezingen om tot uiting. Volgens Dahlgren heeft burgerschap naast een politieke dimensie ook een sociaal-culturele dimensie. Onder burgerschap verstaan we hier de uiteenlopende vormen van politieke, culturele en maatschappelijke participatie. Het gaat, met andere woorden, om de processen waarin mensen vormgeven aan de samenleving, en waardoor zij omgekeerd ook gevormd worden als burger. Media kunnen de burger hierbij enerzijds voorzien van informatie die voor maatschappelijke participatie nodig is en anderzijds kunnen ze de burger ook (impliciet) uitnodigen tot participeren. Er wordt alom getwijfeld of regionale media hun bijdrage aan burgerschap en democratie ook daadwerkelijk (kunnen) vervullen. Uit eerder onderzoek, naar de nieuwsvoorziening over lokale en provinciale verkiezingen, blijkt dat regionale nieuwsmedia veelal een institutioneel perspectief hanteren. Dit bleek onder andere uit het eenzijdige brongebruik: er kwamen vooral veel politici aan het woord terwijl het maatschappelijke middenveld en de burger zelf een geringe rol speelden. Ook de Tijdelijke Commissie Innovatie en Toekomst Pers constateert dat de democratische functie van de regionale journalistiek onder druk staat. Daarnaast verkeren regionale media in een moeilijke economische situatie en kampen met name kranten met dalende bereikcijfers en krimpende redacties. Tegen deze achtergrond komen we tot de volgende onderzoeksvraag: In welke mate en op welke manier brengen regionale media in de reguliere nieuwsvoorziening aspecten van burgerschap in beeld? Deelvragen: 1. Hoe vaak komen verschillende (provinciale) onderwerpen in de reguliere nieuwsvoorziening van regionale media aan bod? 2. Welke bronnen worden in de reguliere nieuwsvoorziening van regionale media in welke mate opgevoerd? 3. Welke rol speelt de burger als bron in de reguliere regionale nieuwsvoorziening? Om de bovenstaande vragen te kunnen beantwoorden is een kwantitatieve inhoudsanalyse uitgevoerd op de berichtgeving in de maand januari van 2012, van regionale kranten (om de andere dag, al het regionale nieuws) en regionale omroepen (dagelijkse tv uitzendingen en actualiteitenrubrieken) in de zes noordoostelijke provincies. In totaal werden er zo’n 1600 televisieitems gecodeerd en bijna 2400 krantenartikelen. Naast de onderwerpcategorieën en bronnen werd van ieder bericht o.a het genre gecodeerd. De rol die burgers speelden werd gecategoriseerd als vox pop, toevallige voorbijganger, 110
111
ETMAAL 2013
BOOK OF ABSTRACTS
betrokkene/slachtoffer, participerende burger of vertegenwoordiger van een groep. De belangrijkste resultaten laten zich als volgt samenvatten. Bij beide media blijken kortere genres te overheersen. Slechts een gering percentage van de berichtgeving betreft context en achtergrond. Qua onderwerpen valt op dat bij beide media de categorie veiligheid/criminaliteit/rechtbank het meest voorkomt: zo’n 25% van alle berichten. Daarnaast is er relatief veel aandacht voor cultuur, media en sport (krant: 20%; tv: 25%). Het politieke nieuws (provinciaal en gemeentelijk) vormt slechts 8% van het nieuws bij de kranten en 6% bij de omroep. Buiten verkiezingstijd is er dus betrekkelijk weinig aandacht voor expliciet politieke onderwerpen. Van de opgevoerde bronnen valt wel 20-25% in de categorieën politici en bestuurders zodat de politieke dimensie gedeeltelijk in het overige nieuws naar voren komt. Maatschappelijke organisaties en ondernemers vormen beide 10% van de opgevoerde bronnen. In tegenstelling tot eerder onderzoek blijkt de burger relatief vaak opgevoerd te worden als bron. In de krant is 40% van de bronnen burger en in het televisienieuws maar liefst 50%. In de krant wordt de burger vaker opgevoerd in een actieve rol (nl. 40% participerend of als vertegenwoordiger van een groep). Op televisie komt zo’n actieve rol in 30% van de gevallen voor. De burger treedt echter het meest op als vox pop (krant: 26%; tv: 43%) en in de rol van slachtoffer of betrokkene (gem. 25%) De belangrijkste conclusie is dat de burger een opmars heeft gemaakt in het regionale nieuws: de burger is inmiddels de meest opgevoerde bron. Het nieuws lijkt hiermee minder institutioneel van aard te zijn geworden. Kanttekening is dat de burger meestal geen ‘actieve’ rol speelt, maak vaak een gemakkelijke, passieve rol tot voorbeeld krijgt. Een verder negatief punt is dat er buiten verkiezingen om weinig politiek nieuws gebracht wordt. Voeg daaraan toe dat slechts in geringe mate context en achtergrond wordt geboden, dan dringt zich de conclusie op dat burgerschap maar matig gestimuleerd lijkt te worden door de regionale media.
GATEKEEPER OF CURATOR? DE ROL VAN SOCIALE MEDIA IN DE BERICHTGEVING OVER DE ARABISCHE LENTE IN DE TV-JOURNAALS VAN VRT EN VTM - REBECA DE DOBBELAER, STEVE PAULUSSEN AND PIETER MAESEELE In deze studie gaan we na in welke mate de nieuwsberichtgeving van de mainstream-media over de Arabische Lente blijk geeft van een kentering in de journalistieke praktijken van nieuwsselectie. Waren er in de berichtgeving indicaties van een verschuiving van de ‘journalist als gatekeeper’ naar de ‘journalist als curator’? Om deze vraag te beantwoorden, opteerden we voor een multimethodische analyse van het televisienieuws van de Vlaamse openbare omroep VRT en de commerciële omroep VTM. In een eerste fase voerden we een kwantitatieve inhoudsanalyse uit op alle nieuwsitems over de Arabische Lente in de zevenuurjournaals van de zenders één (VRT) en VTM, die werden uitgezonden van 1 januari tot 31 maart 2011. In totaal gaat het om 238 nieuwsitems (VRT: 152; VTM: 86). De inhoudsanalyse concentreerde zich op het bronnengebruik in de nieuwsitems. In een tweede fase voerden we een diepgaandere analyse uit op die nieuwsitems waarin gebruik gemaakt werd van sociale media als nieuwsbron. We bestudeerden de manier waarop de journalisten de sociale media in hun reportages verwerkten en de betrouwbaarheid en authenticiteit ervan duidden. Tot slot voerden we diepte-interviews met de buitenlandjournalisten op de nieuwsredacties van VRT en VTM: per redactie werden twee interviews afgenomen. We 111
112
ETMAAL 2013
BOOK OF ABSTRACTS
peilden naar hun percepties over het belang en de problemen van sociale media als nieuwsbron in de journalistieke berichtgeving over de Arabische Lente. Uit de resultaten van het onderzoek blijkt dat de berichtgeving over de Arabische Lente nog steeds in ruime mate bepaald werd door de klassieke nieuwsbronnen: de wereldnieuwsagentschappen (vnl. Reuters en APTN) en buitenlandse media (vnl. BBC). Wanneer user-generated content getoond wordt (vnl. beelden van YouTube), gaat het bijna uitsluitend om fragmenten die door de nieuwsagentschappen werden aangeleverd. Uit de kwalitatieve analyse en diepte-interviews blijkt dat twijfels over de authenticiteit en betrouwbaarheid de buitenlandjournalisten er nog vaak van weerhouden om zelf sociale media als nieuwsbron te verwerken in hun reportages. De bevindingen van ons onderzoek nuanceren de idee dat sociale media leiden tot fundamentele veranderingen in de journalistieke praktijken van nieuwsselectie. Althans voor wat de televisiejournalistiek betreft, zien we dat de wereldnieuwsagentschappen en internationale media dominant blijven als nieuwsbron voor de buitenlandse berichtgeving en dat klassieke routines en mechanismen van ‘gatekeeping’ ook het gebruik van sociale media als nieuwsbron blijven bepalen.
Session #2 Thursday 7 February 15:15 – 16:15 hrs. Chair: Jeroen de Keyser Location: Shanghai (M2-12) English session
DIGITAL AMBULANCE CHASERS, BLOGGING VETS AND NON-STOP COMMENTERS. HYPERLOCAL NEWS IN PRACTICE - PIET BAKKER Hyperlocal media have developed rapidly over the last years. Different models emerged in the USA (The Examiner, Outside-In, Patch) and the UK (Redcliffe, 2012). In Europe myheimat.de was launched in 2005 in Germany; the Belgium hyperlocal Belang van Limburg websites started in 2009; in the Netherland Dichtbij.nl launched in 2010. These models target specific local areas and rely (partly) on volunteers for their content. Compared to previous local online initiatives, the new models are more commercial and usually launched by incumbent media companies, while they have a professional staff and cover large areas by building chains of local operations that share the same technical and organizational infrastructure. Examiner (244 markets in the US) is owned by Denver media-tycoon Philip Anschutz. Patch (800 US communities) is owned by AOL since 2009. Belang van Limburg (all 44 municipalities in the Belgian province Limburg) is part of Concentra, a major Belgium media group. Myheimat.de (covering all of Germany) cooperates with more than a dozen regional publishers (WAZ, Madsack, Neue Presse, etc.). Dichtbij.nl (websites for all Dutch communities) was launched by TMG, controlling 40% of the newspaper market. These websites are no grassroots initiatives (Bowman & Willis, 2003; Gillmor, 2004); but part of major media companies. Trying to engage citizens to participate has been tried often in the last decade. Where some have succeeded (CNN’s iReport, Bild Zeitung’s Leserreporter), others failed. US citizen platform Backfence 112
113
ETMAAL 2013
BOOK OF ABSTRACTS
shut down; Dutch citizen video platform Skoeps closed within one year. Crucial in the success or failure of these platforms is participation. If the number of contributions is too low, there will not be enough content, little visitors and little revenue. In this paper we ask how much hyperlocal platforms rely on amateur contributions (compared to staff-produced and aggregated content), how they engage with volunteers, who the contributors are and why these contributors do participate. Theoretically we draw upon newsroom sociology studies (Blumler, 1969; Breed, 1955; Crouse, 1973; Gans, 1979; Sigal, 1973; Tuchman, 1973) as we argue that organizing a group of people to produce (journalistic) content is best described and explained by these studies. ‘Virtual’ newsrooms, however, do differ from physical newsrooms. Contact between contributors is almost absent; there is usually no subediting or other communication between staff members and volunteers; legal protection is lacking; the professional status of the volunteer is unclear while there are no financial incentives. We concentrate on the differences and similarities of physical and virtual newsroom to explain why people choose to contribute. We also draw upon studies (Costera Meijer, 2012; D'heer & Paulussen, 2012a; 2012b; Redcliffe, 2012) that analysed hyperlocal sites in the Netherlands, Belgium and the UK. During two weeks in 2012 we analysed four editions of Dichtbij.nl in large population (150,000 – 200,000 inhabitants) areas. These websites all cover several municipalities and between 15 and 30 different towns and villages. We analysed the source of all content (volunteers, staff, aggregation), the number and background of the contributors, the number of contributions per volunteer and the subject of their contributions. We also participated in focus group discussions of contributors, and interviewed community managers. Two of the four examined sites only use aggregated content, usually from other local media. In the two remaining editions a quarter of the content is provided by volunteers. A third part is aggregated or comes from press agencies while the remaining articles are staff-written. This last category consists mainly of accidents. Volunteers write about their organization (sports, culture), their businesses (vets, ceramists) or take pictures. Politics and economics are hardly covered. There is a difference between the motives for people to participate in the first place and the reasons to carry on participating. People usually join because they want to raise awareness for a specific subject, or promote a business or an organization. Also more personal reasons (job- or hobby related) can be found. People carry on contributing because of personal contacts with staff and audience members. A few volunteers provide the bulk of the contributions; a large part only participates once or twice. Within the group of commenters the skewed distribution is even more apparent. The five most prolific commenters (with over 500 comments) are responsible for the bulk of all comments. Writers and commenters are distinct groups: writers don’t comment, commenters don’t write articles. The ideal of citizen journalism (a lively, relevant and diverse offering of local journalistic content) is still far from being reached although the number of contributors is still growing. 113
114
ETMAAL 2013
BOOK OF ABSTRACTS
NEWS ECOSYSTEM COMPARED IN TWO DUTCH CITIES - MARCO VAN KERKHOVEN AND PIET BAKKER In thirty years regional newspapers in the Netherlands have seen their circulation drop from 2.7 million in 1980 to 1.8 million in 2012. In some regions daily newspapers have been withdrawn from the market completely. Consequently, from a civic society point of view, the question has been raised what would happen with a regional news ecosystem – defined as all available and interacting media platforms in a given area - if public affairs news was no longer covered by daily newspapers. There is mounting evidence that the daily newspaper is the leading news source in a regional news ecosystem. This might not be surprising taken that a regional newspaper has its roots in the community and is still reasonably well staffed compared to other local media. The question to what extend regional dailies play a vital role in informing audiences about policy and politics is of particular interest. This is especially relevant in the context of the present debate around new business models for regional newspapers and the heavily debated suggestion that regional publishers should receive public financial support. National, regional and local public broadcasters are after all state funded or subsidized based on the idea that the people should have access to independent news sources. Some authors (Gilmore, Bowman & Willis, Jarvis) have suggested that the decline of traditional media like newspapers will be ‘compensated’ by new online initiatives, often with a (hyper) local identity. For three years in a row we analysed the news ecosystem of the so-called no-paper city of Almere. We conclude that, although there are many online news initiatives, only a few independent news organisations persist. Online news blogs and social media are often aggregators without professional staff. Original reporting is limited and public affairs are covered infrequently. Online channels are largely dependent on news from free sheets and the public broadcasters, that in return rely heavily on the municipality and official institutions as their main sources. In this paper we also analysed public affair news in a comparable metropolitan area in the Netherlands, to explore if there are measureable differences in frequency and content that can be explained by the availability of a regional newspaper. We performed content analyses on all news platforms available in an area with a regional newspaper in Tilburg - and an area without a regional newspaper - Almere. We selected a week before the Dutch national elections of September 2012 as our first research period. Elections being a critical event, we expected to see relatively high public affairs news intensity. The second research period was scheduled for the second week afters the elections, a week of which we expected it to be a regular news week. First results show a critical difference in public affairs news frequency between the city with and without a regional newspaper. The city with a regional newspaper produces almost twice as much (90 per cent) original (staff production) public affairs news reports a week (n=86 for two weeks), and an average of 7 stories a day. The six media in Tilburg (newspaper, free weeklies, local and regional TV, websites) apparently produce more (original) public affairs local news than the eight media in 114
115
ETMAAL 2013
BOOK OF ABSTRACTS
Almere (free papers, local and regional TV, websites). Another noticeable finding is that almost 30 per cent of the public affairs news is produced by the regional daily newspaper. Follow up stories, however, are scares, as are interviews in stories. An important source of news is the municipality in both cities. We conclude that in this case the regional daily newspaper produces the majority of the public news stories in Tilburg while Almere lags behind because of an absence of a local paper. Also the diversty in news seems to be bigger at the regional daily. Independent local or hyper local news media have not yet emerged on a substantial scale in both areas.
TRANSPARENCY: WHAT NEWSROOMS (COULD) DO. RESULTS OF AN ONLINE SURVEY AMONG EUROPEAN JOURNALISTS - HUUB EVERS AND HARMEN GROENHART Increasingly, accountability and transparency in journalism is suggested to improve media quality. The main argument is to be found in the assumed power of the profession. Transparency in journalism can be understood as “making public the traditionally private factors that influence the creation of news” (Allen, 2008, p. 323), and this openness may enable the public to value assertions and call the media to account. Other authors suggest effects on trust (De Haan, 2011; Meier & Reimer, 2011; Singer, 2006) or preventing state regulation (Laitila, 1995). Although journalism studies literature offers plenty assumptions on this issue, empirical research on desirability and impact of media accountability practices in journalism is unsatisfying. Because the notion of accountability is closely related to the tension between self-regulation and state regulation, we expect to find valuable insights by international comparative research, between countries form different media systems (Hallin & Mancini, 2004). This paper presents the results of the international and EU funded research project MediAcT, and aims at exploring and understanding instruments, mechanisms, and attitudes towards media accountability and media transparency. The MediAcT consortium conducted a quantitative online survey with Likert-scale items among journalists (N=1764) in 14 European and Arab countries. This paper highlights the findings of a comparative analysis of one aspect in the survey; perceptions of transparency. The central question is: What value do European journalists attach to transparency and how do they perceive their newsroom’s transparency practices? In the analysis we concentrated on the tension between values and practices, and tested hypotheses about different media segments, job positions, age and countries. For the country comparison we calculated centred means for every country on every item (cf. Hanitschz et al., 2011). Generally, we found that journalists attach much value to transparency, claiming their news organization should publish mission statements, codes of ethics and the like. However, we also found remarkable discrepancy between these high values and their generally lower perception of actual practices in the newsroom. We found this discrepancy both related to organizational transparency, audience interaction and newsroom criticism. Specified into media segmentation, we found only a small ‘public institution effect’; public television journalists adhere significantly more value to publishing codes of ethics than newspaper journalists. However, we found a convincing ‘commercial company effect’, as news agency journalists significantly emphasize the organization’s commitment to high standards more than other media segments. Related to job position, we found that editors115
116
ETMAAL 2013
BOOK OF ABSTRACTS
in-chief were less enthusiastic about publishing mission statements and ownership information than reporters and freelancers. Contrastingly, editors-in-chief were significantly more positive on audience interaction than freelancers and reporters, which may be due to the fact that they, and not editors-in-chief, finally have to deal with the audience in daily practice. Related to age, young journalists attach more value to modern forms of audience interaction, like direct communication via social media and online participatory news production. With respect to country comparison, it was hard to find meaningful geographical or cultural clusters. Most countries showed a similar discrepancy between desired norms and actual practices of transparency. Some individual countries showed remarkably distinctive values. Italian journalists seem to have a higher sensitivity when it comes to disclosing information of ownership and business or political affiliation, potentially related to the comprehensive influence of the Berlusconi administration. Dutch journalists attribute relatively little value to responsiveness to the audience, which may be explained as a rudiment of the pillarized society that once guaranteed strong audience loyalty. As these pillars decayed, Dutch journalists gained sensitivity towards the audience, but maybe not yet as much as in other countries. On the level of audience participation, Jordan journalists show highest believes in citizens’ contribution to journalism, which may be related to the Arab spring, where citizen’s voice had significant impact on governmental regimes in both Jordan and Tunis. These results suggest that European journalists share a common ideology on norms of transparency, and all witness a certain discrepancy between norms and practices. There seems no reason for media system thinking, although country specific circumstances clearly resonate in attitudes towards media accountability. In-group differences however, hint towards suggestions for improvement in media accountability, especially when it comes to awareness and beliefs of managers versus operational journalists and young versus older journalist. Journalists seem to shift responsibility for media accountability towards other levels in the news organization, both top down as well as bottom up.
FACING MEDIATED DISTANT SUFFERING: A FIRST STEP TOWARDS AN EMPIRICAL FRAMEWORK JOHANNES VON ENGELHARDT AND JEROEN JANSZ Most of today’s suffering caused by violent conflict and other humanitarian disasters takes place in countries of the so-called developing world. In this configuration, what Western publics know and how they think about large-scale humanitarian disaster is almost exclusively derived – directly or indirectly – from media accounts. It is therefore not surprising that a substantial amount of academic interest has been directed at Western media representations of large-scale human misery in developing countries. At the same time, a considerable body of literature has accumulated during the previous decade which reflects on what it can mean for media audiences to face distant humanitarian disaster, typically building on the concept of cosmopolitanism. However, these theoretical writings have not yet been matched with substantial empirical investigations on how audiences make sense of, and respond to, representations of humanitarian disaster. As noted by Höijer and others, the current literature includes “few empirical studies of audiences’ reactions to and interpretations of the media exposure of distant suffering” (Höijer, 2004, p. 513). This current situation has allowed important questions on Western spectatorship of distant suffering to remain largely a topic of theoretical reflection. A point in case here is the notion that Western audiences have grown emotionally and morally numb 116
117
ETMAAL 2013
BOOK OF ABSTRACTS
towards victims of distant suffering. This “Compassion Fatigue” thesis, while persistently appearing in academic and popular discourse, has never been adequately addressed (let alone confirmed) empirically. This theoretical paper argues that in order to encourage more concerted empirical studies which will advance our understanding of what it means for Western audiences to be confronted with representations of large scale human misery, it is necessary for media studies to reach out to other academic disciplines. Based on a review of the media studies literature on distant suffering, we aim to highlight how theoretical concepts and empirical findings from social and moral psychology speak directly to the ongoing discussion. We thus intend to show how such an effort can provide fresh ideas and renewed impetus to the field of media studies, and in particular to the rich theoretical body of literature on Western media audiences faced with visuals and narratives of humanitarian disaster in the Global South. Our discussion, then, is structured around what we identify as three core themes in the current debate: distance, factuality and scale. For each theme, we aim to invite into the debate new questions, concepts and findings from outside of media studies. First, different conceptualizations of distance have played a dominant role in the media studies literature on distant suffering and experimental studies in psychology have indeed shown how are more likely to relate to and help those that we feel “close” to. At the same time, other forms of distance might significantly problematize the spectator-sufferer relationship. For example, drawing on studies in moral psychology, we show how a lack of “experiental overlap” between those who are suffering and those who are watching could prevent empathetic responses of Western audiences visà-vis distant suffering. From here, we also argue for the necessity to more rigorously distinguish between empathy (as feeling what the suffering other is feeling) and compassion (as feeling for the suffering other) in the field of media studies. Second, regarding factuality, we contend that there is a constant danger of representations of distant suffering to be perceived as possessing a limited degree of actuality. Among others, we integrate findings from research on the “Just World Thesis” on widespread and deep-rooted yet implicit beliefs in the world as a fundamentally fair and just place. This body of literature, which has of yet not been linked to the media studies literature on distant suffering, suggests that audiences might safeguard “just beliefs” by subtly pushing distant humanitarian disaster into the realm of the fictional. Third, regarding scale, moral psychology studies on a “Collapse of Compassion” have repeatedly shown a fundamental human limitation regarding moral-emotional engagement with the suffering of groups of people. This body of research shows how scale matters in that individuals respond considerably less to the suffering of (even small) groups than to the suffering of an individual. This phenomenon then, we argue, should inform future studies on audiences confronted with humanitarian disaster, and potentially modulate overly optimistic assumptions about the potential of representations of mass suffering to cultivate a moral sensibility towards the distant other.
117
118
ETMAAL 2013
BOOK OF ABSTRACTS
Session #3 Thursday 7 February 16:45 – 17:45 hrs. Chair: Steve Paulussen Location: Shanghai (M2-12) English/Dutch session
POLITICI AAN TAFEL - DE INVLOED VAN PROGRAMMAFORMATS OP POLITIEKE INTERVIEWS TIJDENS VERKIEZINGEN - BIRTE SCHOHAUS AND MARCEL BROERSMA Het politieke interview op televisie is de laatste decennia fundamenteel van karakter veranderd. Waar eerder het één-op-één gesprek in journaals en actualiteitenrubrieken maatgevend was, schuiven politici nu steeds vaker aan bij veel bekeken talkshows en infotainmentprogramma’s. Daar worden ze geconfronteerd met een setting waarin onderwerpen snel wisselen, het gesprekstempo hoog is en meerdere interviewers, tafelheren of gasten zich in de conversatie mengen. In het publieke debat wordt vaak gesuggereerd dat deze ‘spannende’ televisie goed is voor de kijkcijfers, maar ten koste gaat van inhoudelijke informatie over politieke besluitvorming. De invloed van programmaformats op dit spanningsveld is in Nederland nog nauwelijks in kaart gebracht. In dit paper wordt onderzocht in hoeverre het format van deze programma’s de interactie tussen politici en journalisten bepaalt en daarmee de beeldvorming over politiek beïnvloedt. Tijdens de campagne voor de gemeenteraadsverkiezingen die plaatsvonden op 3 maart 2010 zijn drie actualiteitenprogramma’s met ver uiteenlopende formats (NOVA, Pauw en Witteman en De Wereld Draait Door) geanalyseerd. Van elk programma is een sample van de twee weken voorafgaand aan de verkiezingen gekozen, alsmede de week na de verkiezingen zodat ook nabeschouwingen mee konden worden genomen in de analyse. De programma’s zijn geanalyseerd met behulp van een inhouds- en beeldanalyse. Naast de lengte van de respectievelijke uitzending werd het aantal items met politieke interviews en de lengte ervan gemeten. De onderwerpen van de interviews zijn geanalyseerd aan de hand van een topic-lijst van zestien onderwerpen. Hierdoor kon de pluriformiteit van onderwerpen binnen een uitzending, maar ook binnen een programma met de andere programma’s worden vergeleken. Binnen de items is het aantal vragen en interrupties gemeten, als ook de spreektijd van interviewer en politicus. Doordat ook het gebruik van filmfragmenten werd geanalyseerd, kan een beeld worden gegeven van de indeling en opbouw van de items. Voorts is het soort vragen in drie categorieën (interpreterend vs. neutraal, persoonlijk vs. inhoudelijk en meegaand vs. confronterend) ingedeeld op een Likertschaal van 1 tot 5. Deze aspecten geven uitsluitsel over de opbouw de interviews en hoe deze samenhangen met het format, en daarmee over het beeld van politiek in deze veelbekeken programma’s. In eerder onderzoek is aandacht besteed aan de rol van de journalist, personalisering van de politicus en het ontstaan van medialogica. Dit gebeurde voornamelijk in het kader van verkiezingsonderzoek of vanuit de taalwetenschap, en was veelal gebaseerd op framingtheorieën of conversation analysis (Clayman 2007 en 2002; Brants en Van Praag, 2005). Volgens aanhangers van de zogenoemde ‘mediamalaise-hypothese’ zou televisie de burger nog slechts een illusie van politieke informatie 118
119
ETMAAL 2013
BOOK OF ABSTRACTS
verschaffen omdat politiek debat en informatie plaatsmaken voor entertainment en sensatie. Anderen stellen dat nieuwsmedia door medialogica overgaan tot interpreterende vormen van journalistiek waarin de politieke inhoud ondergeschikt is aan analyses van politieke strijd (Norris, 2000; Van Zoonen, 2004; Postman, 1985; Hallin, 1992). Dit onderzoek toont aan dat niet zozeer de inhoud, maar voor een groot deel de vorm en setting van de formats de beeldvorming bepalen. De interviewstijl van de presentatoren blijkt daarbij onderdeel te zijn van de formats. De strikte, snelle indeling van DWDD en de grotendeels confronterende en vaak interpreterende vragen geven de indruk van oppervlakkigheid, terwijl bij P&W de setting van een gespreksronde de ‘gezelligheid’ van het programma benadrukt en de presentatoren hun interviewstijl meer aan de opbouw van de uitzending aanpassen. Hoe korter het item hoe interpreterender de vragen om sneller tot een antwoord te komen. Het format van NOVA is het duidelijkst toegespitst op het informeren van de kijker zonder ‘leuke’ elementen die daarvan afleiden. Gezien het hoge gehalte van inhoudelijke en confronterende vragen, kan gesteld worden dat bij NOVA het scherpste inhoudelijke debat is gevoerd. Bij DWDD bleek het format de grootste stempel te drukken op hetgeen de kijker te zien krijgt. De vaste rubrieken bevorderen de herkenbaarheid van het format. Ook de lengte van de items, waarin meestal maar één onderwerp wordt besproken, is strikt afgemeten. Bij NOVA en P&W is deze indeling losser. De lengte van de items wordt door de gasten en onderwerpen bepaald, waarbij ook vaak meerdere onderwerpen in één item aan bod komen. Deze zijn voornamelijk gericht op en bepaald door het nieuws en de actualiteit. Door aandacht te besteden aan de context van de interviews en de beïnvloedende factoren binnen een televisieprogramma voegt dit onderzoek een belangrijk aspect toe aan bestaande studies. De karakteristieke kenmerken van het format van de programma’s zijn, zoals uit dit onderzoek blijkt, van grote invloed op de manier van interviewen en de relatie tussen journalist en politicus en daarmee op het beeld dat de kijker krijgt.
IDEOLOGISCH PLURALISME IN VLAAMSE NIEUWSMEDIA: DE BELGISCHE REGERINGSFORMATIE 2010-2011 - DANIËLLE RAEIJMAEKERS AND PIETER MAESEELE De voorbije jaren is de wereldbekendheid van België aanzienlijk gestegen. Redenen hiervoor zijn geen wafels van topkwaliteit of de expansie van een schaars abdijbier, wel de vorming van de federale regering. Wanneer begin 2010 tijdens de regering Leterme II de onderhandelingen over een zesde staatshervorming spaak lopen, worden er vervroegde verkiezingen uitgeschreven. Zo kan met de steun van de burger een nieuwe regering de onderhandelingen terug aanvatten. De verkiezingen op 13 juni 2010 zorgen echter niet voor het heropleven van de Belgische politiek, integendeel, voor een verlamming. Maar liefst 541 dagen duurt het voordat de nieuwe regering Di Rupo I op 6 december 2011 de eed aflegt en de zesde staatshervorming een feit wordt. Een absoluut wereldrecord. In deze context van diepgaande politieke crisis is het meer dan ooit van belang dat de burgers van democratische samenlevingen geïnformeerd worden over de alternatieve analyses en beleidsprogramma’s die op het spel staan. Gedurende deze 541 dagen wordt de regeringsformatie 119
120
ETMAAL 2013
BOOK OF ABSTRACTS
op de voet gevolgd door de binnenlandse nieuwsmedia. Deze paper onderzoekt welke rol deze nieuwsmedia spelen in het licht van deze democratische uitdaging, door te evalueren in welke mate nieuwsmedia bijdragen aan een ideologisch-pluralistisch debat. Internationaal en nationaal onderzoek naar pluralisme in de pers wordt vooral gekenmerkt door kwantitatieve inhoudsanalyses naar de mate van “partisan media bias” in ontzuil(en)de/commerciële mediasystemen. Dit wordt geoperationaliseerd als de verschillende niveaus van aandacht voor en de evaluatie van specifieke politici of politieke partijen door specifieke media (Biltereyst & Van Gompel 1999; Cuyt & De Swert 2000; D’Alessio & Allen 2000; Distelmans 1999; Groseclose & Milyo 2005; Hopmann, Van Aelst & Legnante 2011; Junqué de Fortuny, De Smedt, Martensa et al. 2012; Larcinese, Puglisi, Snyder 2011; Van Aelst 2007; Van Gompel & Biltereyst 1997). Deze paper wil deze – theoretisch en empirische – focus op politiek, in termen van politici en politieke partijen, overstijgen ten voordele van een focus op het politieke: de ideologische (ie axiologische, normatieve en beleidsmatige) assumpties en voorkeuren vis-à-vis een gegeven politiek-maatschappelijke orde (Carvalho 2007; Mouffe 2005). De vraag stelt zich vervolgens hoe, voorbij een focus op partijpolitieke voorkeur, ideologische voorkeuren en assumpties kunnen worden blootgelegd. In dit opzicht is er recent, in aanvulling op verschillende spraakmakende onderzoeken naar de rol van ideologische factoren in de nieuwsrepresentatie van sociale, economische en politieke kwesties de voorbije decennia (Hall, Critcher, Jefferson et al. 1978; Fairclough 1995; Van den Berg & van der Veer 1989; van Dijk 1991), een hernieuwde belangstelling in media studies ontstaan met kritische discoursanalyse als theoretische en empirische benadering (Carvalho 2007; Maeseele 2009; Phelan 2007a; Thetela 2001). Wegens hun verschillende ideologische achtergrond worden ten eerste de twee zogenaamde kwaliteitskranten De Morgen en De Standaard geselecteerd en de meest verkopende (populaire) krant Het Laatste Nieuws. Daarnaast wordt ook De Wereld Morgen meegenomen. In tegenstelling tot de drie kranten, maakt De Wereld Morgen geen deel uit van een commerciële mediagroep. Als alternatief medium onderscheidt het zich expliciet van traditionele media op zowel operationeel als inhoudelijk vlak. Ten tweede selecteren we enkel de editorialen van elke krant. Editorialen worden algemeen beschouwd als de enige plaats waar expliciet standpunten door een krant worden ingenomen en aspecten worden aangeklaagd, waardoor die als indicatief voor de ideologische cultuur en specifieke politiek-ideologische positionering doorgaan (Fowler, 1999). Een derde selectie slaat op vier kritische discoursmomenten: (i) de verkennende gesprekken van informateur De Wever, (ii) de onderhandelingen door preformateur Di Rupo, (iii) de nota van bemiddelaar Vande Lanotte, en (iv) de onderhandelingen over het sociaaleconomische luik. In de analyse onderzoeken we op basis van een kritische discoursanalyse op deze vier crisismomenten van de formatie welke ideologische voorkeuren en assumpties kunnen worden onderscheiden in de editorialiserende berichtgeving van vier Vlaamse nieuwsmedia, met een focus op de communautaire en sociaaleconomische breuklijn. De resultaten wijzen op een minimale en sterke mate van ideologisch pluralisme tussen respectievelijk de kranten onderling en tussen de kranten en het alternatief online nieuwsmedium. De analyse legt verder twee ideologische culturen bloot die respectievelijk gekenmerkt worden door depolitiserende en politiserende discursieve 120
121
ETMAAL 2013
BOOK OF ABSTRACTS
strategieën. Deze resultaten lijken die school van politieke filosofen te bevestigen die ons tijdperk als postpolitiek en postdemocratisch beschrijven en de ontzuiling van het krantenlandschap blijkt niet tot meer, maar tot minder, ideologisch pluralisme geleid te hebben waardoor de waakhondfunctie als zodanig aan sterke beperkingen onderhevig is.
WHO HAS GOT THE POWER? A SOCIAL NETWORK APPROACH TO THE RELATIONSHIP BETWEEN POLITICIANS AND JOURNALISTS - ANNEMARIE WALTER AND WILLEM ROBERT VAN HAGE A central question in communication science and since recently political science is who has the power in the interdependent relationship between the news media and politics, i.e. who is leading whom? (e.g. Strömback and Nord 2006; Tresch 2009; Brants et al. 2010). Regardless the numerous studies devoted to the power play of these two institutions, limited attention has been paid to the interpersonal relationship between politicians and journalists (see for exceptions Van Aelst et al. 2010; Van Aelst et al. 2008; Aalberg and Strömback 2011; Ross 2010). Research examining the interactions between these two groups of professionals is primarily based on elite interviews or survey data containing perceived power perceptions and reported details on the frequency, nature and initiator of contact. Irrespective the general limitations of survey data, it prevents studying power fully as a relational concept, i.e. power as a property of an interaction varying from relation to relation depending on a specific moment in time and a specific domain. The field is not only in need of more theoretical refection on what the concept of power entails (see Van Aelst 2008), but the concept should also be studied empirically in accordance to its definition. This not only means methods matching theory, but also a stimulation of the development of theory on this symbiotic relationship. Social Network Analysis is such method that allows us to model power relations. In this study we will make a first attempt by using Social Network Analysis techniques to map and explain the interpersonal relationships between politicians and journalists on the micro-blogging and online social network tool Twitter during and after the 2012 Dutch Parliamentary Election Campaign. As source-journalists relationships are always encased with secrecy, we will look at their online interactions. The friendships formed between politicians and journalists on Twitter are not random and Twitter is an increasingly important news source for journalists. We will attempt to answer the central question who has the power in several ways. First of all, we will map the online network of relationships on Twitter and use various measures to determine the most powerful actors and examine who initiates contact. Second, we will make use of Exponential Random Graph Models, or p* models, to explain which relationships are formed and interactions take place and which are/do not. This research makes a threefold contribution. First of all, it contributes to the field of studies that examines the interdependent relationship between politics and the media, in particular the interpersonal relationships. Second, this article demonstrates what Social Network Analysis techniques can mean for this field. Finally, it contributes to academic research on Twitter that is still in its infancy.
121
122
ETMAAL 2013
BOOK OF ABSTRACTS
GEREPETEERDE TONEELSTUKJES OF GEÏMPROVISEERDE VERKIEZINGSDEBATTEN? EEN ONDERZOEK NAAR DE ONDERHANDELINGEN TUSSEN POLITICI EN MEDIA TIJDENS DE NEDERLANDSE VERKIEZINGSCAMPAGNE VAN 2012 - YAEL DE HAAN, HARMEN GROENHART, AD VAN LIEMPT, NIEK HIETBRINK, BERT MATEBOER, ANOUK KRAGTWIJK, FREEK DIEPHUIS, TIJMEN DE GROEN, JIMMY LEENDERS, ERIK WOLTERS AND ARJANNE ALEMAN Weinig journalisten zullen ontkennen dat er wordt onderhandeld met nieuwsbronnen over de aard en voorwaarden van hun mediaoptredens. Het publiek krijgt daar sporadisch iets van mee, bijvoorbeeld toen bleek dat Freek de Jonge als gast bij 1 voor de verkiezingen werd afgebeld, omdat de VVD niet met de Jonge aan tafel wilde zitten. Voor buitenstaanders blijft het doorgaans echter gissen naar de werkelijke aard en omvang van dergelijk overleg, hetgeen voor sommige critici reden is om te pleiten voor meer transparantie op dat vlak (Dersjant et al., 2012). Naast dit democratische motief werpt het overleg met politici vooral nieuwe vragen op over het competentieprofiel van de journalist. Hoe veeleisend zijn bronnen over hun mediaoptredens en hoe gaat een journalist daarmee om? Hoewel de strijd tussen interviewer en geïnterviewde al in 1979 is aangekaart als een complexe relatie waarbij de bron dikwijls de macht heeft (Gans, 1979), lijkt de complexiteit alleen maar te zijn toegenomen (Prenger & Van Vree, 2004). De veronderstelde toename van voorlichters en communicatieadviseurs (Prenger et al., 2011) zal het belang van onderhandelingsvaardigheid vermoedelijk eerder toe dan af nemen. De centrale vraag is: Hoe verlopen onderhandelingen tussen media en politici voorafgaand aan de verkiezingsdebatten voor de verkiezingen 12 september 2012? Met deze studie sluiten we aan op eerdere publicaties over onderhandelingen tussen bron en journalist (Bakker, van Liempt, Broertjes, Smit, 2011) en leggen we de focus op de relatie tussen media en politici in verkiezingstijd. Methode Er is een kwalitatief onderzoek verricht bij prominente kranten en televisieprogramma’s, zoals de Volkskrant, NRC Handelsblad, Trouw, NOS, RTL Nieuws, Een Vandaag, Nieuwsuur, 1 voor de Verkiezingen en De Wereld Draait Door. Voorafgaand aan de verkiezingen, tussen 12 augustus en 12 september, zijn de overlegprocedures tussen journalisten en politici van binnenuit geanalyseerd door dagelijks vergaderingen bij te wonen op de redacties en interviews te houden met politieke chefs en redacteuren. Na de verkiezingen zijn gesprekken gevoerd met de politieke partijen over hun ervaring met overlegprocedures. Resultaten Dit onderzoek laat zien dat onderhandelingen tussen politici en media over campagneoptredens een lange aanloop hebben. Hoewel voor het publiek de campagne omstreeks eind augustus van start ging, werden de eerste belronden en afspraken al gemaakt op de dag dat het Catshuisoverleg strandde in mei 2012. Politici zijn op zoek naar hoe ze hun informatie via de media kunnen bereiken en televisieprogramma’s en kranten zijn met elkaar in hevige concurrentie om de juiste kandidaat op de juiste plek en tijdstip te krijgen. De bevindingen verschillen sterk per mediumtype. Bij omroepjournalistiek spelen vooral de onderwerpen, formats, en overige gasten een rol. We zien dat de onderhandelingen voornamelijk tussen de grootste partijen en populairste televisieprogramma’s plaatsvinden. De kleinere 122
123
ETMAAL 2013
BOOK OF ABSTRACTS
mediapartijen kozen er eerder voor om de campagne te verslaan zonder interviews met politici of unieke formats te bedenken om zo de dominante positie van grote mediaorganisaties te ontlopen. Onder de grote partijen is niet zozeer het onderwerp een punt van onderhandeling als wel wie, wanneer en waar aan tafel mocht schuiven. Zo wilde mevrouw Schippers niet in debat gaan met een andere politicus bij een optreden in Nieuwsuur. Roemer wilde bij De Wereld Draait Door niet dat Peter R. de Vries tegenover hem zou zitten. Bij dagbladen spelen verschillende vormen van autorisatie een grote rol. Hoewel journalisten geen moeite hebben en eerder belang bij hebben als politici interviews nalopen op feitelijke onjuistheden, wordt er door politici ook geprobeerd om invloed uit te oefenen op inhoudelijke componenten en de volgorde van de onderwerpen in het artikel. Zo wilde Slob niet dat het homo-standpunt van de CU aan het eind van het interview werd behandeld. Conclusie Hoewel politici veel eisen stellen en deze soms ook worden gehonoreerd, voelen journalisten zich niet beknot in het overleg met politici. Journalistieke keuzes en argumenten spelen nog een grote rol en daar waar er wel van te voren onderwerpen werden toebedeeld en het aantal spreekminuten is vastgesteld, wordt het publiek steeds vaker hierover geïnformeerd, zoals tijdens het RTL Premiers debat. Vergeleken met het onderzoek van Bakker et al. (2011) vinden wij minder macht bij de bronnen, de politici. De verklaring ligt in het feit dat politici tijdens verkiezingstijd veel free publicity nodig hebben om de kiezer te bereiken. Hoewel veel critici, zoals voormalig hoofdredacteur van NOVA, Carel Kuyl, spreken van “gerepeteerde toneelstukjes”, laat dit onderzoek echter zien dat die toneelstukjes slechts ten dele worden gerepeteerd, waardoor er nog nodige ruimte is voor improvisatie.
Session #4 Friday 8 February 09:00 – 10:00 hrs. Chair: Jeroen de Keyser Location: Shanghai (M2-12) English/Dutch session
NEOLIBERALISM AND THE MARKETING OF MISERY: NEWS FROM CHILE’S “LARGEST GHETTO” ISABEL AWAD In journalism studies, neoliberalism has largely been discussed in economic terms, to describe either the business conditions under which journalists work (e.g. Fenton, 2011; Hallin, 2008) or the news coverage of economic liberalization (e.g., Horner, 2008; Pheland 2007). However, in the literature on neoliberalism, scholars have increasingly called for a broader understanding of this concept. According to Brown (2003, 7), “[n]eo-liberal rationality […] is not only or even primarily focused on the economy; rather it involves extending and disseminating market values to all institutions and social action, even as the market itself remains a distinctive player.” Significantly, this broader approach to neoliberalism may help identify spaces for the “re-appropriation of ‘market’ techniques 123
124
ETMAAL 2013
BOOK OF ABSTRACTS
of government” for non-neoliberal and even anti-neoliberal purposes (Ferguson, 2009, 174). This study is one of the first to apply this broader approach to neoliberalism to journalism studies. Its aim was to investigate to what extent traditionally voiceless and disempowered groups in neoliberal societies employ typically neoliberal strategies to raise their voice and regain power. Specifically, the study looks at how members of a marginalized, pauperized community utilize the news media to demand governmental policies and funding that improve their living conditions. The site for this study was Chile, a country often mentioned as a paradigmatic case of neoliberalism (e.g., Harvey, 2005; Posner, 2008; Burdick et. al, 2009). The empirical research focused on Bajos de Mena, a neighborhood at the outskirts of Santiago, labeled in the media as “Chile’s largest ghetto.” The result of neoliberal urban policies, Bajos de Mena is particularly suitable for this study. More than 120 thousand people live here in social housing built in the 1990s and 2000s. Isolated from most social services and from the rest of Santiago, Bajos de Mena neighbors face some of the most pressing problems in Chilean society, including extreme poverty, delinquency, drug traffic, social segregation and inequality. In the words of one of the journalists interviewed for this study, this neighborhood “has become an icon of the flaws of neoliberalism.” Case study, method and results In early 2011, the community leaders studied in this research started a campaign demanding from the government to demolish all apartments blocks in Bajos de Mena and to pay current owners the money to buy a better house or apartment elsewhere. In October 2012, the Chilean ministry of Housing announced a plan to comply with this request. Although the community leaders say that they will fight until the last block is demolished and all residents move to a better place, they saw the government’s announcement as a success. A multi-methodological approach is used to examine the role of the news media in this success. Three months of ethnographic research in Bajos de Mena, as well in-depth interviews with community leaders were conducted to observe and understand their concerns, organizational strategies and relationship with local authorities and with the news media. These data were complemented with in-depth interviews with journalists who have covered Bajos de Mena in highprofile national media, as well as with a qualitative analysis of their news stories. The analysis points to a significant convergence of interests of journalists and the Bajos de Mena movement. The movement’s leaders effectively use on- and off-line mobilization tools to “marketeer” (and sometimes consciously overplay) their misery in ways that call the media’s attention and force authorities to react. Journalists are offered not only dramatic (and thus selling) stories, but also opportunities to talk about the social inequalities that since 2011 are a top priority in Chile’s public opinion (Latinobarometro, 2011). As one of interviewed journalists explained, Bajos de Mena appears repeatedly in Chilean news because it enables journalists to “embody these neoliberal problems”. That the tools of marketing are prevalent resources in neoliberal societies in general and in the case of Chile in particular is not new. Paley (2001) described Chile as a “marketing democracy,” precisely 124
125
ETMAAL 2013
BOOK OF ABSTRACTS
because its political system is shaped by free market economics and promoted through marketing techniques. What is new in the present study, however, is that it shifts the focus away from government and businesses targeting foreign investors and citizen consumers. While Paley (2001) focuses on marketing strategies that conceal (and thus legitimize) social inequalities, the case of Bajos de Mena shows that this kind of strategies are appropriated by marginalized citizens, who publicize their misery to fight against social inequality. In doing that, the paper contributes to a more complex understanding of neoliberalism in journalism studies.
KLIMAATVERANDERING ALS DEFINITIESTRIJD: EEN ONDERZOEK NAAR DE BEELDVORMING DOOR MAATSCHAPPELIJKE ACTOREN - YVES PEPERMANS AND PIETER MAESEELE In dit artikel gaan we na hoe maatschappelijke actoren klimaatverandering framen en hoe deze frames zich tegenover elkaar verhouden tijdens de klimaattoppen van de Verenigde Naties. Daarnaast onderzoeken we aan de hand van diepte-interviews hoe actoren van de klimaatbeweging en de antagonistische tegenbeweging deze frames trachten te verspreiden in het publieke debat. Uit de resultaten blijkt dat voor elk frame van de klimaatbeweging dat klimaatverandering problematiseert, een tegengesteld frame van de tegenbeweging terug te vinden is dat klimaatverandering deproblematiseert. Daarnaast vinden we tussen beide bewegingen een uiteenlopende verhouding tot de media en het publieke debat terug.
MEDIATING TRAVEL IN ELITE AND POPULAR NEWSPAPERS: MEDIA INSPIRED IMAGINATION OF TURKEY - NILYUFER HAMID-TURKSOY AND GISELINDE KUIPERS The majority of journalism research focuses on traditional “hard news” stories covering political and economic developments. The social and cultural role of “soft news” (e.g., lifestyle, art, travel) has largely been neglected by scholars, since it addresses readers as consumers rather than citizens and has little political authority in contemporary capitalist societies (Hanusch, 2012; Hanitzsch, 2007). As a consequence, the institutional, structural, regulatory and ethical dimensions of hard news journalism are well documented, whereas there is little research on soft news journalism: lifestyle, entertainment, arts, and, the main theme of this article travel. As Janssen, Verboord and Kuipers (2011) argue in their study of art journalism, “soft news” journalists function as gatekeepers: they publicly confirm or reject which forms of leisure are legitimate for specific readerships. Travel journalists similarly filter, frame and judge travel destinations of interest to particular groups of travel consumers. We define travel journalism as a practice aiming to present a destination in order to inspire and encourage home audiences to travel (abroad). Scholars have pointed to several features of this increasingly prominent genre. Travel journalism mediates experience of foreign countries, generating images of travel destinations and the “Others” who live there, and often evaluating the cultures and peoples it covers (Hanusch, 2010). Travel journalism shows how cultural and political relationships among nations are embedded in media texts and provides insight into processes of “Othering” and power relations between nations and cultures (Fürsich, 2002; Fürsich & Kavoori, 2001). Therefore, while seemingly “soft”, travel journalism has implications for citizenship: it provides information and cultural frames about “Others”, and can confirm stereotypes or contribute to transcultural encounters (McGaurr, 2010; Santos, 2004). This genre therefore not only informs but potentially transforms societies, cultures, and groups of people. As such, scholars argue that travel journalism represents both new and previously established structures of meaning, ideology and 125
126
ETMAAL 2013
BOOK OF ABSTRACTS
power (Cocking, 2009; Spurr, 1993). In this article, we focus on travel coverage about Turkey: a country with a multi faceted, contested image. Turkey is a prime tourist destination for Brits. According to the Turkish Ministry of Culture and Tourism, Britain sends the third highest number of visitors to Turkey, after Germany and Russia. In 2011, Turkey received almost 3 million British tourists. However, in political news, Turkey often appears as politically and culturally contested: the exotic Muslim Other, simultaneously inside and outside Europe, a poor, uneducated, underdeveloped nation on Europe’s doorstep (Strasser, 2008). This multi-faceted character of Turkey allows for many possible news “angles” or “frames”, also in travel journalism. We present a comprehensive analysis of 99 travel features from three British broadsheet (Guardian, Financial Times, Daily Telegraph) and three tabloid (The Sun, Daily Mail, Mirror) newspapers, over a five-year period (2005-2010). The British newspaper landscape is characterized by a strong class divide: broadsheet for the élites, tabloids for the masses. Contrasting and comparing two mediums, thus, allows us to gauge the relation between journalistic form and content, and broader institutional settings and target audience. These mediums have salient institutional settings and language, as well as, distinctive consumption styles about travel can provide alternative understanding to the perception of travel journalism practice. Drawing on work relating to travel journalism, we ask how travel stories facilitate the construction of meanings and manufacture a travel destination for the newspaper’s intended reader. We found considerable differences in genre, themes, narrative, language and visual between both types of newspapers. The tabloid press, largely targets lowermiddle and working class readers, largely focusing on all-inclusive mass tourism stories, while broadsheet newspapers criticise such “cheap, packaged bombardment” and introduce an élite construct of the “authentic” and tailor-made Turkish experience, high on cultural capital and explicitly defining class. Nevertheless, our overall results illustrate a number of commonalities: a first person perspective, flowery language, and the absence of criticism or the local’s perspective suggests that travel journalism is dictated by a wider economic logic. As they has a mutually dependent relationship with the travel industry, for advertisement revenues. This obliges the journalists to write more positive stories about Turkey, whilst any negative experience are absent from the travel features.
UIT ONDERZOEK BLIJKT. JOURNALISTIEKE OMGANG MET WETENSCHAPPELIJKE INFORMATIE CHRIS VAN DER HEIJDEN In afgelopen jaren heeft het imago van de wetenschap enorme schade opgelopen. Vooral de Stapelaffaire heeft alom alarmbellen doen rinkelen, met als gevolg commissies, onderzoeksrapporten, aanbevelingen en politieke vragen. Hierbij is de blik voornamelijk gericht op wetenschap en wetenschappers. Maar er is nog een andere kant aan de zaak – een kant die op het Europees continent nog maar zelden wordt bekeken maar wel eens van meer belang zou kunnen zijn dan veelal wordt verondersteld: de journalistieke. In recent onderzoek in de Engelstalige en vooral Amerikaanse wereld (Delgado 2009, Barron en Brown 2012) wordt geconcludeerd dat de uitkomst van veel wetenschappelijk onderzoek vervormd in de openbaarheid komt. Die vervor-ming wordt bewerkstelligd door alle betrokken partijen. Om te beginnen zijn er de communicatiemedewerkers van wetenschappelijke instelling en tijdschriften die in hun persbericht de krenten uit de pap halen. Vervolgens worden die kren-ten in nogal wat 126
127
ETMAAL 2013
BOOK OF ABSTRACTS
gevallen door iets te enthousiaste, op aandacht beluste of net iets gemakkelijk te verleiden wetenschappers via een interview onvoldoende genuanceerd, om niet te zeggen bevestigd – in de loop der tijd zijn hiervan tal van pijnlijke voorbeelden te vinden, met als een van de meest opmerkelijke de Nederlandse affaire Buck in 1990. Uiteindelijk zijn het journalisten die in hun berichtgeving het onderzoek reduceren tot een of twee oneliners. Aldus gaat het communicatieproces van wetenschap naar journalistiek als door een (sensa-tie)trechter: een over het algemeen complex, veelomvattend en genuanceerd onderzoek met voorzichtige hypotheses worden gereduceerd tot één of enkele boutades die vooral een goh-effect bij de lezer, kijker of luisteraar ten doel lijken te hebben. Als het goed is, is voorzichtigheid journalisten tot een tweede natuur. Er is geen hoor zonder wederhoor en principieel is het wantrouwen tegenover alle infor-matie die uit de tweede hand komt. Ook zou je verwachten dat het toenemend aantal academici onder de journalisten deze behoedzaamheid nog eens doet toe-nemen – nog afgezien van het feit dat zij bij uitstek in staat zouden moeten zijn academisch onderzoek te beoordelen. Tot slot kan het haast niet anders, zo zou je denken, dat alle ophef van de afgelopen jaren en de rapporten die verschenen zijn naar aanleiding van onder meer de affaires Stapel en Smeesters (Universiteit Tilburg 2011, Erasmus Universiteit 2012) journalisten extra voorzichtig heeft gemaakt als het wetenschappelijk onderzoek betreft. Maar kloppen deze veron-derstellingen? Wie in Lexis-Nexis de zinsnede ‘uit onderzoek blijkt’ intikt, zal zien dat deze tus-sen eind september 2011 en eind september 2012 zo’n 3000 keer werd gebruikt, verreweg het meest door het Dagblad van het Noorden (134) en enkele andere regionale bladen (De Gelderlander, Brabants Dagblad, De Twentse Courant Tu-bantia) maar ook vaak door De Telegraaf (91 keer), NRC Handelsblad (65), de Volkskrant (58) en Trouw (57). NRC.Next staat met 30 keer onderaan de lijst. Op de keper beschouwd is de zinsnede ‘uit onderzoek blijkt’ opmerkelijk. Nie-mand zou zoiets ooit schrijven als het om een mededeling van een burgemeester of politicus gaat. Zij hebben meningen. Wetenschappers zijn daarboven blijkbaar verheven. Journalisten schrijven immers niet: in een onderzoek van X wordt ge-steld dat Y, nee, uit onderzoek blijkt. Temeer gezien de eerder genoemde beroe-ringen in wetenschapsland zou je verwachten dat een dergelijke zin alleen nog met de grootste voorzichtigheid geschreven wordt. Dat is dus niet zo. Vandaar de dubbele vraag van dit onderzoek: Waarop is de zinsnede ‘uit onderzoek blijkt’ gebaseerd: persberichten, in-terviews, berichtgeving door derden (kranten, tijdschriften) of kijken jour-nalisten ook het onderzoek zelf in? Verder: wordt het bericht gecontroleerd? Hoe en bij wie? Is er sprake van wederhoor? In dit paper worden deze vragen onderzocht aan de hand van artikelen die in genoemde periode met de zinsnede ‘uit onderzoek blijkt’ verschenen zijn in (de gewone krant, niet de wetenschapsbijlagen van) Het Dagblad van het Noorden, De Telegraaf en het NRC Handelsblad. De uitkomsten worden afgezet tegenover de bevindingen uit de eerder genoemde Engelstalige literatuur en het weinige dat hierover dat in Nederland (Van Maanen 2012) naar voren is gebracht.
127
128
ETMAAL 2013
BOOK OF ABSTRACTS
Session #5 Friday 8 February 10:15 – 11:15 hrs. Chair: Pytrik Schafraad Location: Shanghai (M2-12) Dutch session
MEER LOKAAL NIEUWSAANBOD, MEER VAN HETZELFDE NIEUWS - QUINT KIK, PIET BAKKER, LAURA BUIJS AND JUDITH KATZ Lokale en regionale Nederlandse nieuwsmedia staan onder druk. De regionale dagbladoplage is gedaald, van 1,9 miljoen in 2000 naar 1,4 miljoen in 2011 en de concurrentie neemt steeds meer af. In 2000 woonde 35% van de Nederlanders in een gemeente waar twee of meer kranten verschenen (Bakker, De Ridder & Schönbach, 2010), in 2011 was dat nog maar 15%. Hier tegenover staat echter een stijging van het aantal online aanbieders van lokaal nieuws. Door de komst van deze online initiatieven, neemt de druk op traditionele lokale nieuwsmedia nog verder toe. Er is echter nog weinig bekend over het aantal nieuwe online initiatieven en de aard van deze initiatieven. Het nieuwe lokale medialandschap is nog nooit volledig in kaart gebracht. Dit onderzoek zal inzicht bieden in het totale aantal lokale nieuwstitels op gemeenteniveau. Om dit inzicht te bieden, wordt zowel gekeken naar het traditionele nieuwsaanbod (dagbladen, nieuwsbladen, huis-aan-huisbladen, radio en televisie) als naar het online nieuwsaanbod (websites van traditionele nieuwsmedia, zelfstandige online nieuwstitels en de online ‘doorgeefluiken’ van nieuws dat elders is geplaatst, de zogeheten aggregatiesites). Er wordt antwoord gegeven op de vraag uit hoeveel lokale nieuwstitels een inwoner van een gemeente in Nederland gemiddeld kan kiezen. Regionale verschillen worden eveneens in kaart gebracht. Naast het kwantitatieve inzicht zal worden getoetst worden of er een relatie is tussen de aanwezigheid van traditionele lokale nieuwsmedia en zelfstandige online nieuwstitels. Een aanbodsinventarisatie van alle 418 gemeenten is op deze schaal binnen Nederland uniek, ook in het buitenland ontbreekt vergelijkbaar onderzoek. Eerder onderzoek (Bakker, 2002; Verschuren & Memelink, 1989; Vergeer, 2006) binnen Nederland concentreerde zich op traditionele media. In 2010 is internet wel in een aanbodsinventarisatie betrokken. Bakker, De Ridder en Schönbach (2010) inventariseerden naast lokale print- en audiovisuele media ook het lokale aanbod op internet, maar dit onderzoek beperkte zich tot een gewogen steekproef voor het aanbod in twintig gemeenten. Het huidige onderzoek heeft plaatsgevonden in een periode van vier maanden (juni tot en met september 2012). Voor de inventarisatie van het aantal regionale en lokale printtitels is de database Handboek Nederlandse Pers van Nijgh geraadpleegd, cijfers over dagbladen zijn afkomstig van Cebuco en gegevens over omroepvergunningen komen van het Commissariaat voor de Media. Voor de inventarisatie van digitale lokale nieuwstitels is online onderzoek uitgevoerd. In eerste instantie zijn websites gelieerd aan traditionele nieuwsmedia geïnventariseerd. Daarnaast zijn elk van de 418 gemeentenamen afzonderlijk ingevoerd in Google, gevolgd door het keyword ‘nieuws’. Vervolgens zijn de eerste drie Google-pagina’s geanalyseerd. Aan de hand van een coderingsinstructie zijn de 128
129
ETMAAL 2013
BOOK OF ABSTRACTS
gevonden resultaten onderverdeeld in drie deelcategorieën: (1) zelfstandige online nieuwstitels, (2) zelfstandige online nieuwstitels, met deels geaggregeerd nieuws en tot slot (3) pure aggregatiewebsites. Een Nederlandse gemeente telt in 2012 gemiddeld 29 lokaal gerichte nieuwstitels (print, omroep en online). Deze lokaal gerichte nieuwstitels bestaan uit 10 traditionele nieuwstitels (dagbladen, nieuwsbladen, huis-aan-huisbladen, radio en televisie) en ruim 18 online nieuwstitels. Van deze online nieuwstitels zijn meer dan 6 titels gelieerd aan traditionele nieuwstitels, ruim 10 titels zijn online doorgeefluiken van nieuws dat elders is geplaatst (aggregatiesites) en 1 à 2 titels zijn zelfstandig. Het totale online nieuwsaanbod per gemeente doet op het eerste gezicht dus vermoeden dat het aanbod vrij ruim is, terwijl relatief gezien het aanbod van lokaal eigen nieuws van zelfstandig opererende journalisten vrij beperkt is. Daarnaast blijkt dat de grootte van een gemeente wat betreft inwonertal van invloed is op de aanwezigheid van traditionele lokale nieuwsmedia, de online afgeleiden en de zelfstandige online initiatieven. Hoe groter de gemeente, hoe groter het lokale nieuwsaanbod. De aanwezigheid van traditionele lokale nieuwsmedia blijkt eveneens een positieve invloed te hebben op het aantal zelfstandige online initiatieven. In gemeenten met een bovengemiddeld aantal traditionele nieuwstitels zijn meer zelfstandige online nieuwstitels aanwezig, dan in gemeenten met minder nieuwstitels. Dit zou kunnen impliceren dat traditionele offline nieuwsmedia dienen als informatiebron voor de zelfstandige journalisten. Kortom, de voornaamste conclusies van het onderzoek luiden dat er in korte tijd meer lokale nieuwskanalen zijn bijgekomen, maar dat een inwoner van een gemeente daarmee niet automatisch beschikt over meer oorspronkelijk nieuws. Daarbij geldt bovendien dat aanwezigheid van traditionele nieuwsmedia een positieve invloed blijkt te hebben op het ontstaan van zelfstandige online nieuwstitels.
VAN BOETE NAAR BELONING. PRIORITEITSVERSCHUIVING IN PUBLIEKSVERANTWOORDING DOOR DE JOURNALISTIEK - HARMEN GROENHART Diverse journalisten pleiten al enige tijd voor meer openheid in hun vak (Luyendijk, 2006; Smith, 2008) en aanpassingen in de beroepscodes (GvH, 2008; RvdJ, 2007) laten zien dat dit pleidooi ook langzaam in het vak institutionaliseert. De journalistiek deelt dit met andere sectoren, getuige verantwoordingsarrangementen in het openbaar bestuur (Bovens & ’t Hart, 2005; RoB, 2012), de geneeskunde (Timmermans, 2005) of de rechtspraak (Raad voor de Rechtspraak, 2007). Critici beweren echter dat het huidige niveau van verantwoording door journalisten nog onvoldoende is (Bardoel, 2005; Rouw & Schillemans, 2003; Harinxma thoe Slooten, 2007; Vermeulen, 2012), hetgeen impliceert dat er kenbare maatstaven zijn voor de aard en functie van verantwoording in de journalistiek. Bovendien verwacht het werkveld dat publieksverantwoording specifieke eisen gaat stellen aan de vaardigheden van de toekomstige journalist (Drok, 2011), hetgeen gevolgen heeft voor het beroepsonderwijs. Deze vaardigheden en de aard en functie van verantwoording zijn vooralsnog onvoldoende gespecificeerd, onder meer vanwege een uiteenlopend en soms tegenstrijdig 129
130
ETMAAL 2013
BOOK OF ABSTRACTS
begrippenkader. Deze paper levert een bijdrage aan de exploratie en beoordeling van publieksverantwoording en richt zich op de vraag: Op welke wijze leggen journalisten verantwoording af aan hun publiek en wat betekent dat voor de Nederlandse journalistiek? Het onderzoeksobject is afgebakend aan de hand van instrumentele (Bertrand, 2000) en procesmatige definities (Pritchard) van publieksverantwoording, alsmede op basis van het onderscheid in frames (Bardoel & d’Haenens, 2004; McQuail, 2003). De gehanteerde methodiek van deze paper – gebaseerd op verschillende deelstudies van een proefschrift naar publieksverantwoording door de journalistiek in Nederland – is meervoudig en bestaat uit literatuurstudie, kwantitatieve analyse van websites van nieuwsmedia (n=75), kwalitatieve en kwantitatieve argumentatie-analyse van hoofdredactionele verklaringen (N=131), diepte-interviews met journalistieke professionals (N=47) en focusgroepen met nieuwsgebruikers (N=33). De argumentatie-analyse, de interviews en focusgroepen concentreerden zich rond vier mediumtitels: dagblad Trouw, het Eindhovens Dagblad, het actualiteitenprogramma Nieuwsuur en Omroep Gelderland. Deze paper richt zich specifiek op het ontwaren van verschillende opvattingen van journalisten over publieksverantwoording en verschuivingen in de dominantie van die opvattingen. In het uiteenlopende onderzoeksmateriaal kwam vooral samenhang naar voren in beweringen over verantwoordelijkheid, transparantie en de huidige staat van de journalistiek. Ordening van deze beweringen leidde tot vijf verschillende discoursen van publieksverantwoording. Het verbeteringdoor-sanctiediscours vat journalistieke verantwoordelijkheid op als iets dat moet worden afgedwongen door andere partijen en transparantie als middel om kwalijke praktijken aan het licht te brengen. In dit discours is de journalistiek in staat van verval en moet zij disciplinair verbeterd worden. Het zelflegitimeringdiscours ziet journalistieke verantwoordelijkheid als een privilege dat verdiend moet worden en transparantie als een middel om professionele kwaliteit en integriteit te tonen. In dit discours wordt de journalistiek bedreigd in haar bestaansrecht door de aanwas van andersoortige zenders zoals adverteerders en voorlichters, en moet zij zich overtuigend aan de samenleving presenteren. Het dialoog-als-functiediscours ziet publieksverantwoording als een onderdeel van de journalistieke platformfunctie. Transparantie betreft het openlijk bediscussiëren en beoordelen van het journalistieke proces met het oog op aansluiting met de samenleving. De journalistiek experimenteert met de toepassingen van deze dialoog, die met de komst van gedigitaliseerde communicatie steeds meer uitingsvormen kent. Het marktprofijtdiscours betreft het verzilveren van reeds bestaande kwaliteit die overeenkomt met verwachtingen van journalistieke verantwoordelijkheid. Het tonen van uitgangspunten en procedures is een vorm van merkprofilering waarmee nieuwsorganisaties betrouwbaarheid suggereren. De journalistiek is van voldoende kwaliteit, alleen ontbreekt het aan de vaardigheid zichzelf te verkopen. In het professioneleefficiëntiediscours tenslotte, gaat de journalist strategisch in dialoog met publiek en gaat hij open en transparant te werk, hetgeen hem ondersteunt bij het realiseren van zijn taken. De journalistiek bevindt zich in een transitiefase en innoveert door middel van diverse aspecten van transparantie en dialoog. Deze studie wijst op een intensivering van publieksverantwoording in de laatste decennia, aangezien nieuwsmedia zich transparanter profileren en nadrukkelijker interactie zoeken met het publiek. Maar deze intensivering laat zich niet verklaren met de traditioneel normatief georiënteerde literatuur die 130
131
ETMAAL 2013
BOOK OF ABSTRACTS
vooral een sanctiegerichte benadering voorstelt. Waar in het verleden transparantie en verantwoording exclusief werden gezien in de context van maatschappelijke verantwoordelijkheid, vindt er nu, tenminste in de praktijk, een meer zakelijk discours plaats dat met name de commerciële en praktische toepasbaarheid van publieksverantwoording onderzoekt. Voorbeelden daarvan zijn: het creëren van merken rondom individuele journalisten, het vergroten van loyaliteit door toegankelijkheid en druk zetten op weigerachtige nieuwsbronnen. Dergelijke toepassingen impliceren geen fundamentele verandering van het vak, maar vereisen wel nieuwe vaardigheden die getraind en aangestuurd kunnen worden.
“KWESTIE VAN GEZOND VERSTAND!?” EEN STUDIE NAAR DE ZIN EN ONZIN VAN SOCIALE MEDIARICHTLIJNEN BIJ VLAAMSE JOURNALISTEN - HARALD SCHEERLINCK AND MICHAEL OPGENHAFFEN Situering onderzoek Onderzoek in Vlaanderen, Nederland en daarbuiten heeft aangetoond dat meer en meer journalisten gebruik maken van sociale media als Facebook en Twitter om nieuws te researchen, te presenteren en te breken, en om met de nieuwslezer te converseren (e.g. Hermans, Vergeer, Pleijter, 2011; Quadrant Communications, 2012). Ook de nieuwsmedia zelf zien het belang van sociale media in en stimuleren hun journalisten om actief te zijn binnen sociale media en de nieuwslezer meer betrokken te maken. Ze zien echter ook de gevaren in het gebruik van sociale netwerksites door de journalisten. Onderzoek van Lasorsa et al. (2012) heeft immers aangetoond dat journalisten binnen Twitter heel vaak hun eigen mening verkondigen over het nieuws en dit zou de journalistieke objectiviteit in het gedrang brengen. Ook zijn er al verschillende voorbeelden bekend van journalisten die een primeur via Twitter lekten zonder dat hun hoofdredacteur hiervan op de hoogte was. Het is dan ook geen verrassing dat binnen heel wat nieuwsorganisaties de discussie gevoerd wordt over wat eigenlijk mag en niet mag binnen sociale media. Een aantal nieuwsmedia hebben als gevolg hiervan een luikje rond sociale media in hun deontologische codes opgenomen of zelfs een aparte richtlijn opgemaakt, dit om misbruik en discussies achteraf te vermijden. Zo publiceerde Reuters al in maart 2010 specifieke richtlijnen voor het gebruik van sociale netwerksites en dit document begint met: “We want to encourage you to use social media approaches in your journalism but we also need to make sure that you are fully aware of the risks […]” (Reuters, 2010). In de maanden nadien worden ook binnen de redacties van oa. Associated Press, Sky News, BBC, NRC Handelsblad, VRT, VTM, … sociale mediarichtlijnen opgesteld die vertellen hoe je als journalist binnen Facebook en Twitter moet omgaan met primeurs, met onbevestigde berichten, met het retweeten van berichten, met het overnemen van foto’s, enz. Maar niet alleen binnen nieuwsorganisaties wordt er bepaald wat kan en niet kan binnen sociale media. Ook de Raad voor de Journalistiek van de Vlaamse journalistenvereniging schreef in 2012 een nieuwe richtlijn uit die journalisten vertelt hoe ze zich moeten gedragen binnen sociale media. Deze richtlijn omvat bijvoorbeeld een richtlijn die het gebruik van informatie en beelden afkomstig van sociale netwerksites en persoonlijke websites regelt. Onderzoeksvraag Het is duidelijk dat sociale media en journalistiek steeds meer en meer verweven wordt. Het mag dan ook geen verrassing zijn dat mediaorganisaties sociale mediarichtlijnen hebben uitgeschreven om 131
132
ETMAAL 2013
BOOK OF ABSTRACTS
hun journalisten te wijzen op voordelen, maar zeker ook op de nadelen van Facebook en vooral Twitter tijdens het journalistieke productieproces. De vraag rest echter hoe de journalist deze richtlijnen ervaart. Dit onderzoek bevraagt daarom Vlaamse journalisten naar de zin en onzin van richtlijnen die het gebruik van Twitter moet regelen. Methodologie We hebben om bovenstaande onderzoeksvraag te beantwoorden 20 Vlaamse beroepsjournalisten geïnterviewd. Voor het selecteren van respondenten maakten we gebruik van een selecte steekproef, op basis van volgende criteria: graad van activiteit op Twitter (alleen actieve Twitteraars), platform (zie boven), mediabedrijf (Corelio, De Persgroep, Mediafin, Sanoma Belgium, Roularta, Belga, VMMa, …), functie (zeven respondenten hadden een coördinerende functie om eventuele verschillen tussen chefs en redactieleden te identificeren) en geslacht (14 mannen, 6 vrouwen). De vragen gingen over het bestaan van richtlijnen, welke adviezen er precies instaan, of ze deze al dan niet relevant vinden en waarom, hoe ze zelf omgaan met bijvoorbeeld scoops en retweets, wat de sancties zijn bij overtreding, enz. Resultaten Uit de analyses van de interviews blijkt dat de Vlaamse journalist geen heil ziet in het op papier zetten van sociale mediarichtlijnen. Het meest gehoorde argument hiervoor is dat gezond verstand zou moeten volstaan. Sommige journalisten ervaren de regels zelfs als een inperking van hun individuele vrijheid. Vooral de richtlijnen omtrent specifieke vormvereisten zoals het vermelden van de werkgever in de biografie en/of accountnaam of de verplichting slechts één account te mogen gebruiken stuit op grote weerstand bij de Vlaamse journalisten. Slechts een klein aantal journalisten zien de zin van richtlijnen wel in, maar enkel wanneer deze geformuleerd worden als tips en dus niet als afdwingbare regels. Ook op de richtlijn vanuit de Raad voor de Journalistiek zit de Vlaamse journalist niet te wachten. Slechts een kleine groep vindt het een goed idee, als was het maar om de discussie levendig te houden en er eventueel tips uit te kunnen halen. Tijdens de presentatie gaan we dieper in op de resultaten en de implicaties hiervan voor nieuwsorganisaties.
Session #6 Friday 8 February 11:45 – 12:45 hrs. Chair: Karin Raeymaeckers Location: Shanghai (M2-12) Dutch session
FREELANCERS IN DE VLAAMSE JOURNALISTIEK: VRIJE VOGELS OF VOGELS VOOR DE KAT? KWALITATIEF ONDERZOEK NAAR RANDVOORWAARDEN FREELANCESTATUUT - HEDWIG DE SMAELE AND ROZANE DE COCK Achtergrond Aan de hand van cijfermateriaal uit tweejaarlijkse online surveys wordt regelmatig een profiel 132
133
ETMAAL 2013
BOOK OF ABSTRACTS
geschetst van de populatie Vlaamse beroepsjournalisten met aandacht voor loon, genderverdeling, leeftijd, opleiding, dagelijks takenpakket, competenties, specialisatie en dergelijke meer (zie onder meer Raeymaeckers, Paulussen & De Keyser, 2008; Paulussen & Raeymaeckers, 2010). In deze informatieve Vlaamse maar ook buitenlandse studies wordt echter zelden expliciet gefocust op freelance journalisten (Edstrom & Ladendorf, 2011, p. 1) en de keuze voor een kwalitatieve onderzoeksmethode is nog zeldzamer. Ons onderzoek (lente 2012) wil specifiek de situatie van freelancers in Vlaanderen belichten en laat 22 van hen hierover zelf aan het woord in uitgebreide diepte-interviews. De keuze voor ons specifieke onderzoek bij freelance journalisten is ingegeven door zowel onderzoeksmatige aanwijzigen als door signalen uit het werkveld. Onderzoek binnen de algemene journalistenpopulatie toont immers aan dat een te grote werkdruk, onregelmatige werkuren, avondwerk en weekendwerk een knelpunt vormen binnen het beroep (Paulussen & Raeymaeckers, 2010) en bovendien lopen journalisten een verhoogde kans op burn-out (Teugels, Van Hoof, Mory & De Witte, 2009). Al deze problemen komen nog duidelijker tot uiting binnen de groep freelancers en zij kampen bovendien met een ongunstig werkstatuut met veel minder bescherming en zekerheden dan hun collega’s in vaste loondienst. Bij de starters is het aandeel mensen aan de slag met een freelance statuut niet gering: 10% werkt er voor meerdere opdrachtgevers en 9% voor een vaste opdrachtgever (Paulussen & Raeymaeckers, 2010). Het werken voor één vaste opdrachtgever maar niet in dienstverband, brengt het thema van de schijnzelfstandigheid onder de aandacht. Zeker in economisch ongunstige tijden voelen freelancers de repercussies het eerst (Apache, 02.05.2012). Zweeds onderzoek (Edstrom & Ladendorf, 2011, p.4) wijst nochtans op de voordelen die freelancers ervaren (verworven vrijheid, grotere tevredenheid) en het bestaan van zowel een portfoliomodel (freelance als eigen keuze) als een marginalisatiemodel (freelancen als dwang). Onderzoeksvragen In ons kwalitatief onderzoek willen we vooral peilen naar de randvoorwaarden voor een haalbare freelance journalistiek in Vlaanderen. We formuleren de volgende onderzoeksvragen: OV 1: Wat zijn volgens freelance journalisten in Vlaanderen de randvoorwaarden voor een haalbare freelance journalistiek? a. professionele invulling (nieuwsproduct, opdrachtverwerving, belang sociaal netwerk, werkuren, extra competenties als freelancers versus journalist in loondienst, …) b randvoorwaarden context (tijdsinvulling, combinatie werk-gezin, verloning) c. randvoorwaarden sociodemografische situatie (opleiding, belang specifieke opleiding, gender, leeftijd, belang startkapitaal, vangnet inkomen partner/andere steun, …) d. professionele gratificaties (Waarom ooit gekozen voor freelance statuut? Belang van erkenning binnen journalistiek en bij publiek, …) OV 2: Wat geven Vlaamse freelancers aan als voordelen en nadelen van freelancen? OV 3: Hoe tevreden zijn Vlaamse freelance journalisten over hun werk als freelancer? OV 4: Hoe zien Vlaamse freelance journalisten de toekomst tegemoet? Zien zij zichzelf blijven binnen de mediasector in dit werkstatuut? Methode In de lente van 2012 werd uit de databank met contactgegevens van Vlaamse Beroepsjournalisten 133
134
ETMAAL 2013
BOOK OF ABSTRACTS
(verkregen via de Vlaamse Vereniging voor Journalisten, VVJ) een bestand opgesteld met 71 toevallig geselecteerde freelance journalisten uit de geschreven pers (dagbladen en magazines), online, radio en televisie. Deze journalisten werden telefonisch of via mail gevraagd mee te werken aan ons onderzoek. 22 freelance journalisten werden bereid gevonden mee te werken aan onze studie. Een vooraf opgestelde gespreksgids met centrale thema’s vormde de leidraad doorheen de diepteinterviews. De interviews werden met toestemming van de betrokken journalist (elke interviewee ondertekenende een informed consent formulier) digitaal opgenomen, volledig en letterlijk uitgeschreven en vervolgens gecodeerd (labels per tekstframe) en geanalyseerd. Samenvatting resultaten De meeste geïnterviewde Vlaamse freelance journalisten geven aan dat het voor hen belangrijk is “eigen meester” te zijn van hun tijd én opdrachten. Maar voor deze vrijheid moet ook een prijs betaald worden: freelancers hebben een onzeker bestaan en moeten tijd investeren in zogenaamde “managementstaken”. Vrijheid staat immers niet gelijk met ledigheid: freelance journalisten moeten voortdurend hun eigen PR verzorgen en hun voelsprieten uitsteken op zoek naar nieuwe opdrachten. Randvoorwaarden die zij naar voren schuiven zijn: beschikken over een startkapitaal, kunnen terugvallen op een partner met een vast inkomen, veel sociaal kapitaal verzamelen (netwerk), beroep doen op professionele hulp wat betreft fiscaliteit en beschikken over specifieke extra competenties die loontrekkenden niet nodig hebben (managements- en PR-kwaliteiten). Een unique selling proposition voor jezelf kunnen uitwerken is eveneens cruciaal (specialisatie). Wie doelbewust kiest voor freelance journalistiek is meer tevreden over zijn werksituatie dan wie hier noodgedwongen in belandt. Die eigen vrije keuze blijkt cruciaal in de beleving. Wat door de ene groep als nadeel wordt ervaren (financieel aspect, het door elkaar lopen van werk en privé) blijkt immers door de andere groep net als voordeel aangegeven te worden. Freelancen: geen zwartwitverhaal.
PROFESSIONELE IDENTITEIT VAN JOURNALISTEN WERKZAAM IN VERSCHILLENDE MEDIATYPEN LIESBETH HERMANS Maatschappelijke, economische en technologische veranderingen die hebben plaatsgevonden in de samenleving hebben ook allerlei gevolgen voor de nieuwsvoorziening. De monopolie positie van de journalistiek als verantwoordelijke institutie voor de productie van nieuws is niet meer vanzelfsprekend. Er vindt discussie plaats over de vraag waar de hedendaagse journalistiek staat en hoe de toekomst van de journalistiek eruit gaat zien (zie oa Broersma et. al, 2012; Costera Meijer, 2006; Drok, 2006, Van Ummelen, 2009, special issue TvCW 2011). Auteurs beargumenteren de ontwikkelingen vaak in de context van een tweestrijd, namelijk een strijd tussen de ideologisch/conventionele benadering en de meer economische/commerciële benadering. Sinds het einde van de 20ste eeuw komt er aanvullend aandacht voor een benadering waarbij de aandacht van de journalistiek zou moeten liggen bij het creëren van burgerschap . Waarin er vanuit de behoeften om betrokkenheid bij burgers te stimuleren meer aandacht wordt bepleit voor de agenda van de burger (oa. Drok 2006). Journalisten zijn onderdeel van een journalistieke cultuur. De hier boven kort geschetste veranderingen hebben gevolgen voor de nieuwsorganisaties en communicatiepatronen daarbinnen. Daardoor zullen de ontwikkelingen ook invloed hebben op het socialisatieproces binnen de 134
135
ETMAAL 2013
BOOK OF ABSTRACTS
redactionele omgeving, waardoor het ook weer gevolgen heeft voor hoe journalisten tegen hun beroep aankijken. Journalisten krijgen te maken met een situatie waarin allerlei vanzelfsprekendheden en grenzen verdwijnen. Naast de concrete praktische gevolgen heeft het vervagen van professionele kaders ook gevolgen voor de journalistieke identiteit. Deze identiteit is gebasseerd op journalistieke ideologische uitgangspunten die onder andere zichtbaar worden in de journalistieke waarden en normen welke zijn samen te vatten in een aantal overkoepelende constructen. Het betreft (1) de maatschappelijke functie van journalisten, namelijk de verantwoordelijkheid voor het openbaar maken van infomatie ten behoeve van een goed functionerende democratie, (2) het handhaven van journalistieke waarden zoals betrouwbaarheid en neutraliteit, (3) onafhankelijkheid en autonomie van de journalsitiek, (4) actualiteit, en (5) ethische afwegingen (Deuze, 2005). In dit onderzoek wordt er van uitgegaan dat, vooral de economische veranderingen, niet voor alle mediatypen even grote gevolgen hebben gehad. Regionale dagbladen lijken bijvoorbeeld meer financiele gevolgen, en daardoor bezuinigingen, te ondervinden dan landelijke dagbladen, daarnaast komen de economische gevolgen bij private ondernemingen anders tot uitdrukking dan bij publieke media. Tegen deze, hier kort geschetste, achtergrond, willen we in dit onderzoek kijken in hoeverre er sprake is van een homogene beroepsgroep of dat er verschillen zijn in opvattingen die journalisten hebben over “ideologische” aspecten van hun beroep. Daarbij kijken we naast individuele en professionele kenmerken in het bijzonder naar kenmerken van het mediumtype waar journalsten werkzaam zijn. Data zijn afkomstig van een landelijk online survey gehouden in 2010, onder de leden van de Nederlandse Vereniging van journalisten. Uitnodigingen zijn verstuurd via het e-mail bestand van de NVJ. Als individuele en journalistieke achtergrondkenmerken zijn meegenomen geslacht, leeftijd, professionele opleiding, aard contract. Journalistieke identiteit is bepaald door de begrippen beroepsrolopvattingen, journalistieke waarden, ethische afwegingen, publieksbeeld te meten aan de hand van meerdere stellingen (4punts antwoordschalen). Hiervan zijn gemiddelde berekend voor verdere analyses. Bij de kenmerken van de mediatypen is een onderscheid gemaakt tussen landelijke versus regionale/lokale media en publieke versus private media. (Eerste voorlopige) Resultaten tonen aan dat er verschillen zijn tussen individuele en journalistieke achtergrondkenmerken en bepaalde opvattingen die journalisten hebben over hun beroep. Over het algemeen blijken deze verschillen echter relatief klein te zijn. Ten aanzien van de verschillen tussen de opvattingen van journalisten die werkzaam zijn bij medatypen met specifieke kenmerken zien we dat journalisten werkzaam bij regionale media op meerdere aspecten afwijken van hun collega’s die werkzaam zijn bij landelijke media. Eenzelfde resultaat lijkt ook zichtbaar te zijn voor de journalisten werkzaam bij privaat/publike media. Maar ook hier lijken de verschillen over het algemeen niet zo groot. In het paper wordt uitgebreid ingegaan op de specifieke verschillen die zijn gevonden en hoeverre er sprake is van een patroon. In de conclusies worden de resultaten besproken tegen de achtergrond van de diverse aspecten die onderdeel vormen van de professionele identiteit. In de conclusies worden de resultaten daar waar 135
136
ETMAAL 2013
BOOK OF ABSTRACTS
dat mogelijk is besproken tegen de achtergrond van resultaten van eerder Nederlands onderzoek. Daarbij wordt ingegaan op de discrepantie die er lijkt te bestaan tussen de veranderende werkomgeving van de journalisten en de relatief stabiele ideologische beroepsopvattingen van journalisten.
FOREIGN CORRESPONDENTS IN THE NETHERLANDS - BERNADETTE KESTER For more than 150 years, world news leadership has been exercised by Europe and United States but Europe has been the world news leader at least since 1990 (Tunstall, 2011). Despite its leadership, most research on foreign correspondents concentrates on those based in US (Hess 1996, 2006, Gross and Kopper, 2011), while research in the way that Europe and individual European countries are covered by foreign correspondents is almost non-existent, with notable exceptions [Morrison and Tumber (1985), Hannerz (2004), Mazzoleni and Splendore (2007)]. This will change with the publication of a book (see below), presumably next year. My investigation of ‘Foreign Correspondents in the Netherlands’ is still more or less work in progress (although the actual research has been completed; deadline is December 2012) and part of a larger ECREA-book project, called ‘Foreign Correspondents in Europe’, initiated by Georgis Terzis (Vrije Universiteit Brussel) in 2011. More than 35 researchers from European countries, working in the field of journalism and media & communication studies, are involved in this international comparative research project. Working from a similar questionnaire and in-depth interview questions, every member of the project focuses on the make-up of the foreign European press corps: their numbers, backgrounds, the pressures they are under and their national obsessions. We analyze the different factors affecting the coverage of European correspondents, resources devoted to foreign bureaus, the political affiliation of journalists and media organizations, tensions between parachuted correspondents and old-timers, and public versus commercial media. A detailed breakdown of the key international media players and media sectors follows. Included is also an analysis of the news sources used by of foreign correspondents, the different national institutions and embassies as well as other smaller ‘official sources’ are analyzed together with a number of unofficial ones, such as think tanks, NGOs and lobbies. Reporting by foreign correspondents from The Netherlands is particularly interesting because of the position the Netherlands historically took within the context of international European relations. Recently this position changed. Having a long history of free speech, press and tolerance in general, it subsequently occupied a neutral position during the First World War and after the Second World War it became a so called guiding country, a country with a moral mission so to speak. The Hague became well known for the Peace Palace and the International Criminal Court. This all changed after the elections of 2010 which were very favorable for the nationalist, xenophobic and anti-European party of Geert Wilders (PVV). Without its support on various political issues the government would have fallen much earlier than it finally did. From this historical and current perspective (although recently the political situation has changed again) it is not surprising that The Netherlands have a long tradition in being a guest country for foreign correspondents. In 1925 the Foreign Press Association of The Netherlands was raised and until today is still active. Around hundred correspondents are member of this organization, which is 136
137
ETMAAL 2013
BOOK OF ABSTRACTS
quite a lot for a small country like the Netherlands. No recent research has been done about this organization and its members. Especially today as so many politicians and intellectuals worry about the once so positive connotation The Netherlands had, it is very relevant to investigate who are responsible for the reporting about this country and how this related to the larger context of Europe. In other words, who are these ‘professional strangers’ stationed in Netherlands and in Europe at large and how do they try to make their stories that are clearly important in today’s interconnected world interesting for viewers and readers? That question will be addressed in more detail during the Etmaal conference.
'JOURNALISTIEK IN VLAANDEREN DOOR EEN ROZE BRIL' EEN VERKENNEND ONDERZOEK NAAR GENDERONGELIJKHEID IN DE JOURNALISTIEKE SECTOR EN BEROEPSTEVREDENHEID BIJ VROUWELIJKE JOURNALISTEN - SARA DE VUYST AND KARIN RAEYMAECKERS De laatste jaren is het aantal vrouwelijke journalisten aanzienlijk toegenomen. Daarmee is de kloof tussen mannen en vrouwen die hun intrede maken in het journalistieke veld in Westerse landen bijna volledig weggewerkt (Chambers, Steiner & Fleming, 2004; Demoor, Saeys, De Bleeckere, De Clercq & Reymenants, 2000; Paulussen, Raeymaeckers, De Keyser, Van Leuven, 2010). Sommige auteurs verwachten dat deze kwantitatieve vooruitgang van vrouwen in de journalistiek, er automatisch toe zal leiden dat vrouwelijke journalisten een ‘kritische massa’ vormen, die een tegenwicht kan bieden aan de dominante mannelijke journalistieke organisatiecultuur, routines en nieuwspraktijken (Gallagher & Von Euler, 1995; Mills, 1997; Ross, 2001). Deze assumptie wordt door andere auteurs als te optimistisch ingeschat omdat er geen rekening gehouden wordt met structurele factoren die de genderongelijkheid in stand houden, waardoor de journalistieke organisatiecultuur gekenmerkt wordt door een diepgewortelde verticale en horizontale gendersegregatie (de Bruin & Ross, 2004; Löfgren-Nilsson, 2010; Phalen, 2010; Ziamou, 2000). Dergelijke structurele barrières zijn schijnbaar tegenstrijdig met het hoge niveau van beroepstevredenheid bij vrouwelijke journalisten dat reeds veelvuldig in kwantitatief onderzoek is vastgesteld (Delano, 2003; Deprez, 2010, Johnstone, Slawski & Bowman, 1976; Miller & Miller, 1995, Weaver & Wilhoit, 2007). Het vernieuwende karakter van deze paper bestaat erin de paradox tussen het hoge niveau van genderongelijkheid in de journalistiek en de hoge beroepstevredenheid bij vrouwelijke journalisten meer diepgaand te onderzoeken. Aan de hand van kwalitatieve interviews met 19 vrouwelijke journalisten uit verschillende generaties, mediasectoren en functies brengen we in kaart welke drempels zij ervaren bij de uitoefening van hun beroep, hoe het gesteld is met hun algemene beroepstevredenheid en welke strategieën zij toepassen om met die barrières om te gaan. Tijdens de gesprekken werden alle ‘traditionele’ obstakels bevestigd en kwamen er zelfs bijkomende moeilijkheden naar boven. Bevraagde journalistes verwezen onder andere naar het broos evenwicht tussen journalistiek en een gezin, het negatieve effect van stereotiepe man-vrouwopvattingen op de carrière van vrouwelijke journalisten en de verhoogde werkdruk door de komst van sociale media. Alle respondenten hadden echter in overeenstemming met resultaten van voorafgaand kwantitatief onderzoek, een hoge algemene beroepstevredenheid. Uit de interviews blijkt dat deze paradox kan verklaard worden doordat de respondenten een aantal strategieën aanwenden als buffer om de negatieve impact op hun algemene beroepstevredenheid in 137
138
ETMAAL 2013
BOOK OF ABSTRACTS
te dijken, en zo te verhinderen dat ze hun journalistieke carrière vroegtijdig stopzetten. Eén van die strategieën is het ontkennen van enige vorm van genderongelijkheid in de journalistieke sector. De respondenten die deze strategie toepassen, zijn ervan overtuigd dat er dankzij de toename van het aantal vrouwelijke journalisten de laatste jaren, gelijke condities zijn voor mannen en vrouwen in de journalistiek. Ze wijzen maatregelen of initiatieven die specifiek gericht zijn op het verbeteren van de situatie van vrouwen in de journalistiek af. De ‘laissez-faire’-houding van deze journalistes loopt parallel met de ‘kritische massa’-logica. Opvallend is dat deze respondenten hun redenering ook zelf tegenspreken bij vragen over gendergevoelige deelaspecten van het journalistieke beroep (combinatie werk-gezin, promotiekansen, …), waarbij ze wel voorbeelden van ongelijke behandeling aanhalen. Een andere strategie is het beschouwen van de onregelmatigheid in de job, die de combinatie van journalistiek en kinderen moeilijk maakt, als een voordeel. De onvoorspelbaarheid wordt geassocieerd met andere positieve aspecten zoals vrijheid, flexibiliteit en afwisseling en maakt de job in de ogen van respondenten die deze strategie volgen zo interessant, dat een uitstap geen optie is. Nog andere tactieken die aan bod zullen komen in deze paper hebben te maken met de flexibiliteit op de redactie ten aanzien van journalisten met kinderen, het terugvallen op de steun van een partner, terugschroeven en het gebruik van nieuwe communicatietechnologie. We concluderen dat er ondanks de toename van het aantal vrouwen op redacties in Vlaanderen de laatste jaren, nog verschillende gendergerelateerde barrières aanwezig zijn, maar dat vrouwelijke journalisten heel creatief zijn in de manier waarop ze hiermee omgaan, en zo kunnen vermijden dat deze obstakels interfereren met hun positieve kijk op het journalistieke beroep. Een doctoraatsproject binnen het Center for Journalism Studies (CJS) aan de Universiteit van Gent zal de verschillende hindernissen voor vrouwen in de journalistiek verder onderzoeken, zowel op redacties als binnen journalistieke opleidingen.
Session #7 Friday 8 February 09:00 – 10:00 hrs. Chair: Piet Bakker Location: T3-29 Dutch session
PRIKKELEND TELEVISIENIEUWS EN DE GEVOLGEN VOOR DE KIJKER: OVER AANDACHT VASTHOUDEN, INFORMEREN EN BETROUWBAAR OVERKOMEN - MARISKA KLEEMANS Al sinds jaar en dag hebben nieuwsjournalisten als primaire taak om burgers op een betrouwbare manier te informeren over belangrijke gebeurtenissen in de samenleving. Met name de toenemende concurrentie op de nieuwsmarkt heeft er echter toe geleid dat het voor nieuwsjournalisten ook steeds belangrijker wordt om de aandacht van een zo groot mogelijk publiek te trekken. In zowel publieke discussies als wetenschappelijke studies wordt verondersteld dat deze groeiende strijd om de aandacht van de kijker ervoor heeft gezorgd dat er in het hedendaagse televisienieuws steeds meer prikkelende (‘arousing’) inhouds- en vormgevingskenmerken verschijnen. Deze kenmerken, zoals negatieve onderwerpen en beelden, snelle camerawisselingen en allerlei montagetechnieken, zouden (kortstondig) de aandacht van het publiek trekken, hetgeen ook is bevestigd in diverse 138
139
ETMAAL 2013
BOOK OF ABSTRACTS
studies. Critici bekijken de trend naar steeds prikkelender nieuws met argusogen: zij vrezen dat prikkelend nieuws niet in staat is om de informatieve functie van nieuws te vervullen voor een breed publiek. Centraal in deze studie staat de vraag of dit daadwerkelijk zo is. Nieuws kan alleen de informatieve functie vervullen als het de aandacht trekt en vasthoudt, kijkers informeert, en als kwalitatief betrouwbaar wordt gezien. Aangetoond is dat prikkelend nieuws de aandacht kortstondig trekt, maar nog onbekend is of het ook in staat is om die aandacht vast te houden. Dit is belangrijk om te weten, omdat kijkers alleen geïnformeerd kunnen worden via televisienieuws wanneer zij tijdens het nieuws niet naar de afstandsbediening grijpen, maar het nieuws daadwerkelijk blijven volgen (RQ1). Vervolgens kan de vraag worden gesteld in hoeverre prikkelend nieuws in staat is om kijkers iets te leren. Immers, nieuws kan alleen informatief zijn als mensen er ook iets van onthouden (RQ2). Tot slot is er de vraag naar het vertrouwen dat kijkers hebben in de kwaliteit van prikkelend nieuws. Wanneer dit niet aanwezig is, zullen nieuwsjournalisten er immers niet in slagen om hun informatieve taak te volbrengen (RQ3). In het kader van een promotieonderzoek zijn in de afgelopen jaren diverse onderzoeken uitgevoerd naar prikkelend televisienieuws. Data afkomstig van twee experimenten, een onder hoogopgeleide mannen en vrouwen variërend in leeftijd (N = 288) en een onder hoog- en laagopgeleide jongeren (N = 89), kunnen worden gebruikt om de onderzoeksvragen te beantwoorden. In beide experimenten hebben proefpersonen gekeken naar nieuwsberichten die varieerden in inhoud (negatief versus neutraal) en vormgeving (vlot versus standaard). Proefpersonen hebben vervolgens vragen beantwoord, onder andere met betrekking tot hun voorkeuren, herkenning van nieuws, en ervaren kwaliteit en sensationeel gehalte van de berichten. Ten aanzien van de vraag in hoeverre de aandacht van kijkers kan worden getrokken en vastgehouden door prikkelend nieuws (RQ1), werd in het experiment uitgevoerd onder hoogopgeleiden gevonden dat vooral jongeren een sterke voorkeur hadden voor negatieve inhoud in nieuws. Vlotte vormgeving bleek vooral geliefd bij mannelijke kijkers, terwijl vrouwen het nieuws liever in een standaard vormgeving zagen. Prikkelend nieuws lijkt daarmee niet in staat om bij alle kijkers de aandacht te trekken en vast te houden. Het experiment onder hoog- en laagopgeleide jongeren, dat werd uitgevoerd in het licht van RQ2, toonde aan dat beide groepen weliswaar een voorkeur hadden voor negatief nieuws, maar dat alleen hoogopgeleide jongeren hiervan profiteerden in termen van informatieverwerking. De hoger opgeleiden, en dan vooral de mannen, hadden een betere herkenning van negatief nieuws dan neutraal nieuws, terwijl lager opgeleiden niet beter geïnformeerd werden door negatief nieuws. Dit impliceert dat de veronderstelde kenniskloof tussen hoog- en laagopgeleiden toeneemt door prikkelende inhoud in nieuwsberichten. Het vertrouwen dat kijkers hebben in prikkelend nieuws (RQ3) werd onderzocht onder hoogopgeleiden. Zij evalueerden negatief nieuws en vlot vormgegeven neutraal nieuws als sensationeler. Echter, alleen de negatieve nieuwsberichten werden door jongeren en mensen van middelbare leeftijd ook ervaren als van een iets lagere kwaliteit. Geconcludeerd kan worden dat prikkelend nieuws niet in staat lijkt te zijn om de informatieve functie van nieuws voldoende te vervullen voor een breed publiek. Immers, vooral hoogopgeleide jonge mannen voelen zich tot prikkelend nieuws aangetrokken en herinneren zich ook meer van deze nieuwsberichten. Lager opgeleiden, vrouwen en ouderen ondervinden minder positieve effecten van 139
140
ETMAAL 2013
BOOK OF ABSTRACTS
prikkelend nieuws. Bovendien worden negatief nieuws en tabloid vormgegeven neutraal nieuws als sensationeler bestempeld en blijkt dat naarmate kijkers jonger zijn, zij negatief nieuws vaker zien als nieuws met een iets lagere kwaliteit dan neutraal nieuws. Al met al is het een uitdaging voor journalisten om hun informatieve taak te vervullen in een tijd waarin alles spannender en sneller lijkt te moeten zijn.
EEN KWESTIE VAN VERTROUWEN: DE CONSTRUCTIE VAN DE GELOOFWAARDIGE VERTELLER IN HET JOURNALISTIEKE BOEK - GERARD SMIT Zowel bij lezers als bij journalisten is in Nederland veel belangstelling voor goed geschreven en diepgravende journalistieke verhalen. Die belangstelling blijkt onder andere uit de verkoopcijfers van boeken van gerenommeerde journalisten als Geert Mak en Jeroen Smit, uit de massale toestroom van jonge journalisten bij de eerste twee conferenties over verhalende journalistiek in Nederland, en uit de sinds 2010 jaarlijks uitgereikte MJ Brusseprijs voor het beste journalistieke boek. Net als dat met de opkomst van de literaire journalistiek in de jaren zeventig het geval was, staat ook nu weer het waarheidsgehalte van deze vorm van journalistiek ter discussie. (zie onder andere Metze, 2010). Wat hierbij opvalt is dat die discussie zich beperkt tot de vraag in hoeverre de vermelde feiten verifieerbaar zijn (Kramer, 2007). Wat doorgaans buiten beschouwing blijft, is de vraag in hoeverre de manier van vertellen meespeelt in het oordeel over de geloofwaardigheid. Volgens Lee Gutkind (2012) baseren lezers hun oordeel over een journalistiek verhaal - juist door de onduidelijke relatie tussen tekst en werkelijkheid bij dit genre - mede op het vertrouwen dat ze hebben in de verteller. Hoe geloofwaardiger de verteller, hoe overtuigender het verhaal. Het is voor de journalist dan ook van belang om zichzelf als geloofwaardige verteller te presenteren. De vraag is alleen: hoe doe je dat? Wat maakt je als journalist tot een geloofwaardige verteller? Handboeken over literaire journalistiek (Cheney, 1987; Hart, 2011; Gutkind, 2012; Kramer, 2007; Stewart,1998) geven daar nauwelijks uitsluitsel over. Er wordt wel over het belang van een persoonlijke stem van de verteller gesproken, maar de auteurs gaan niet in op de literaire middelen die je kunt inzetten om de vertellersrol overtuigend in te zetten. Veel verder dan de tip ‘relax en wees jezelf’ (Hart, 2011, p. 73) komt het niet. Anders dan in de journalistieke handboeken wordt in de academische literatuur over verhaalanalyses wel uitvoerig stil gestaan bij de vraag welke factoren bijdragen aan de geloofwaardigheid van de verteller. Er wordt daarbij vooral gelet op de volgende aspecten van de vertellersrol die zijn ontleend aan de literaire kritiek (Booth, 1983), en aan de narratologie (Abott, 2008; Bal 2009; Genette 1972): het karakter van de verteller (voice), de mate waarin de lezer zich kan inleven in de beschreven werkelijkheid (‘narratieve afstand’), het blikveld van waaruit de beschreven werkelijkheid is waargenomen (‘focalisatie’), en de mate waarin de verteller zich rekenschap geeft van zijn eigen positie tegenover de getoonde werkelijkheid (‘zelfreflectie’). In dit onderzoek gebruik ik deze vier kenmerken van de verteller om een duidelijker beeld te krijgen van de manier waarop schrijvers van journalistieke boeken de vertellersrol invullen. De centrale vraag van dit onderzoek is welke verteltechnieken de voor de MJ Brusseprijs van 2012 genomineerde journalisten gebruiken om zichzelf als geloofwaardige verteller te presenteren, en hoe ze dat doen. Deze groep is gekozen omdat het hier specifiek ‘narratieve journalisten’ gaat. Op grond van tekstanalyses geef ik aan hoe de genoemde elementen van voice, focalisatie, afstand, 140
141
ETMAAL 2013
BOOK OF ABSTRACTS
en zelfreflectie, zijn te lokaliseren in de zes genomineerden boeken (Bukman, 2011; Geurts, 2011; Koene, 2011; Kok, 2011; Riemersma, 2011; Soetenhorst, 2011). Vervolgens stel ik vast in hoeverre deze elementen conform de daarvoor in de geraadpleegde literatuur genoemde richtlijnen bijdragen aan de geloofwaardigheid van de verteller. Via diepte-interviews met de auteurs van deze boeken ga ik vervolgens na in hoeverre de journalisten bewust gebruik maken van de verschillende vertelstrategieën om hun geloofwaardigheid als verteller te bevorderen. De voorlopige resultaten wijzen erop dat journalisten bewust een informele toon gebruiken en ernaar te streven hun eigen stem te laten doorklinken in het verhaal (voice). De verhalen zijn zo geschreven dat de lezer de beschreven gebeurtenissen voor zijn eigen ogen ziet ontrollen (geringe narratieve afstand). De meeste journalisten kiezen ervoor het centrum van de waarneming bij zichzelf te leggen (externe focalisatie). Er wordt gebruik gemaakt van zelfreflectie over hun eigen rol als verteller om daarmee het vertrouwen van de lezer te winnen. Dit onderzoek is beperkt in omvang en exploratief van aard. De resultaten moeten niet worden gezien als bewijs voor een theorie over journalistieke geloofwaardigheid, noch als de aanzet voor een richtlijn voor journalistiek schrijven. Het onderzoek is vooral bedoeld om greep te krijgen op een tot nu toe veronachtzaamd aspect in de discussie over de waarheid van journalistieke verhalen, namelijk de geloofwaardigheid van de verteller.
DUTCH JOURNALISM PRACTICES IN THE DIGITAL AGE - BOB NIEMAN, MAX KEMMAN, MARTIJN KLEPPE AND HENRI BEUNDERS With an ever-growing supply of digital and online sources, information to produce news stories seems to be one mouse click away when using these computer-aided research tools. Several scholars point to the impact of digital technologies on journalistic work (Pavlik, 2000; Deuze, 2007; van Dijk, 2004) but also instructional books to do research online have been published widely (van Ess, 2010; Dasselaar & Pleijter, 2010). Furthermore, recently we have seen the establishment of several associations that provide hands-on trainings such as the Dutch Flemish Association for Investigative Journalism (VVOJ) and the Association for Dutch Online Journalists (VOJN). But to what extent do Dutch professional journalists actually use computer-aided research tools? Machill and Beiler (2009) show in a similar research that German journalists use online tools more frequent, but for shorter periods than traditional research tools, eventually concluding these digital tools complement traditional research methods while the telephone remains the most important research tool. Information about the use of computer-aided tools by Dutch journalists is limited. Hermans et al. (2010) state the Internet is mainly used to follow the news, check facts and search for background information. However, both structured research on the use of computer-aided tools by Dutch journalists as well as ethnographic observations of the research process of journalists is scarce (Brants & Vasterman 2010). This paper aims to fill this gap by answering the main research question: To what extent are digital sources incorporated in the working process of Dutch journalists? It gives more detailed information by answering the sub-research questions: 1) To what extent are online databases and search engines used? 2) Which search techniques are applied? 3) Are there differences between the type of medium (television, newspaper, magazine) and journalistic specialism (e.g., politics, economics, sports, arts & culture)? Building upon the methodological approach of Machill and Beiler and similar research on the use of 141
142
ETMAAL 2013
BOOK OF ABSTRACTS
digital sources by academics (Kemman, 2012), we applied a combined methodological approach by conducting an ethnographic study as well as a survey. We first observed 13 news editors from a Dutch daily newspaper, in their search process for information and analysed how they used the Internet to complement their information channels. In order to get more quantitative data, we conducted an online survey amongst a broad group of Dutch journalists, receiving over 300 respondents working at newspapers, television programmes and magazines. In this survey we focused on the online sources journalists use to get their information, how they find this information and what search techniques they apply in doing so. This survey is organized in the framework of the European FP7 research project ‘AXES - Access to Audiovisual Archives project’ (see www.axesproject.eu). Journalists are one of the user groups studied within this project. The first preliminary results of our study show Dutch journalists use relatively few computer-aided research tools. In line with Machill and Beiler, we found the Internet hasn’t radically changed the way journalists gather their information, but has become another channel of information equally important as other channels such as the telephone. When using online sources, two are dominant: Google and Wikipedia. The use of Google is limited to the first results page, and use of other online sources is limited to websites already known to the journalist. In the first phase of the search process, a journalist searches with broad search terms. After choosing an approach to the story, more specific search terms are formulated. Therefore it appears the main question for journalists when using computer-aided research tools is not which search engine should be used, but which search terms are most effective.
HYPERTEXT EN DIRECT SPEECH IN ONLINE NIEUWSARTIKELEN. HET VERGROTEN VAN DE ERVAREN CONTROLE EN INLEVING IN HET VERHAAL - DANIËL VERHEIJ AND LUUK LAGERWERF Sinds de komst van het internet proberen de nieuwsmedia zich te onderscheiden van andere nieuwssites om zo een vaste schare lezers voor zich te winnen. Internet biedt nieuwsorganisaties allerlei nieuwe mogelijkheden die met de traditionele media als televisie, radio en krant niet mogelijk waren. Niet alleen kunnen beeld, geluid en tekst op een nieuwssite gecombineerd worden, het internet biedt daar ook de ruimte toe. De journalist hoeft een nieuwsverhaal niet meer in te korten om de tekst passend te maken: het internet kent geen maximum aantal woorden. De grote vraag is echter of een een bezoeker van een nieuwssite wel zit te wachten op lange teksten. Interactiviteit, hypertext en multimedialiteit worden al sinds het ontstaan van internet geprezen als de toekomst van de journalistiek (Dimitrova, 2010; Paulussen, 2004). Sluiten deze mogelijkheden aan op de nieuwsbehoefte van jongeren (Costera Meijer, 2007)? Zij zien volgens Costera Meijer (2007) graag meerdere lagen in het nieuws. Ze willen het nieuws zien door de ogen van de hoofdrolspelers. Die wens botst met de voorkeur van nieuwssitebezoekers voor korte en overzichtelijke artikelen (Sturgill, Pierce & Wang, 2010). Hypertext kan een oplossing zijn voor dit probleem. Als de kern van het verhaal kort verteld wordt kunnen verschillende bronnen achter links worden verborgen. Geïnteresseerde lezers kunen zich verdiepen in de verschillende kanten van het verhaal, anderen kunnen doorlezen. Om te achterhalen of de wijze waarop de bronnen achter de links worden weergegeven een rol speelt werden ze met quotes of indirect opgevoerd (Semino, Short, & Culpeper, 1997; Ekström, 2006). De probleemstelling luidt: kunnen in hyperlinks weergegeven bronnen de lezer van een online nieuwsverhaal meer 142
143
ETMAAL 2013
BOOK OF ABSTRACTS
waardering en tevredenheid bieden? Een online experiment is uitgevoerd onder 83 respondenten. Twee nieuwsverhaen werden gemanipuleerd. Een verhaal ging over de Amerikaanse presidentskandidaten, met als bronnen oudklasgenoten, kandidaat Rombey, en tegenstander Obama. Het tweede verhaal ging over ProRail. Hier werden bronnen van de VVD, PvdA en SP geciteerd. Van beide verhalen werd een variant zonder, en een met hyperlinks geconstrueerd. Bovendien werden de geciteerde bronnen met directe quotes of indirect geciteerd. Respondenten kregen van beide onderwerpen een verhaal te lezen, waarbij (in)directe rede werd gevarieerd. De aanwezigheid van hyperlinks was een tussenproefpersoonfactor. Afhankelijke variabelen waren ervaren controle, leesplezier, inleving in het verhaal en tevredenheid. Als modererende persoonskenmerken werden daarnaast behoefte aan controle en hypertext comfort gemeten (Berger, 2001; Burger & Cooper, 1979). Uit de resultaten bleek dat hypertext leidde tot verschillen in controle en in het leesplezier: wie een nieuwsbericht met hypertext leest, ervaart meer controle over de tekst. Daarbij gold dat hoe hoger de ervaren controle, hoe meer leesplezier de respondenten hadden. De behoefte aan controle speelde hierbij een modererende rol. Het gebruik van direct speech maakte weinig verschil voor de inleving in het verhaal, of voor de ervaren controle over de tekst. Citeren in directe rede wordt geacht de levendigheid en inleving in een nieuwsverhaal te vergroten (Semino, Short, & Culpeper, 1997), maar dat bleek niet het geval in dit onderzoek. Ervaren controle over de tekst is daarentegen wel een factor die effect heeft op leesplezier. Het maken van nieuwsverhalen waarin verschillende perspectieven in hyperlinks worden opgenomen kan een bijdrage leveren aan het suces van een nieuwswebsite.
Session #8 Friday 8 February 10:15 – 11:15 hrs. Chair: Liesbeth Hermans Location: T3-29 English/Dutch session
SHIFTS IN FLEMISH CULTURAL JOURNALISM: AN ANALYSIS OF HISTORICAL TRANSFORMATIONS IN THE COVERAGE OF CULTURE IN FLEMISH ELITE AND POPULAR NEWSPAPERS (1955-2010) - SIL TAMBUYZER In recent years, there has been an increasing amount of literature on cultural journalism (Hellman & Jaakkola, 2011; Jaakkola, 2010; Jannsen, 1999; Kristensen, 2010; Pasquier, 2010). Yet, in these studies, opinions tend to differ about the status of cultural journalism. On the one hand there are the more pessimistic ‘cultural critics’ who lament the current state of cultural journalism often referring to the marginalized ‘high-brow’ arts, less room for criticism and the rise of commercialization. The ‘cultural optimists’ on the other hand, state that cultural journalism is expanding to a more newsdriven and less aesthetically inspired sort of journalism, with attention for more democratic cultural forms as a result of shifting cultural hierarchies in society. 143
144
ETMAAL 2013
BOOK OF ABSTRACTS
In these studies, authors often interpret the notion of culture differently, ranging from a more narrow culture definition, mainly restricted to ‘the (high) arts’ (e.g. Jannsen) to a broad notion of culture including lifestyle, media, and consumer topics such as food, television, games, journalism (e.g. Kristensen). This is leading to different claims about the current status of cultural journalism in general and is making comparisons between results difficult (Kristensen, 2010). In Flanders, a longitudinal analysis on transformations in cultural journalism is lacking. Therefore, this study will analyze evolutions in Flemish culture coverage, starting from an inclusive broad approach of culture including coverage of arts, culture, media, lifestyle and consumption topics. Hereby, we want to point to changes in the cultural hierarchy and legitimacy of these domains and put the most important transformations in an historical perspective (1955-2010). In order to analyze these historical transformations, we selected three established and long standing Flemish newspapers of which two elite newspapers (the traditionally catholic newspaper De Standaard and the traditionally socialist newspaper De Morgen) and one popular newspaper (the traditionally liberal newspaper Het Laatste Nieuws). Our extensive content analysis, which is the first of this range in Flanders, covers 4 time frames (1955-1975-1995-2010) and involves more than 30 variables of which only a selection will be used in this paper. The total number of newspapers analyzed was 104, and the total of items in the dataset is N=4.552. Preliminary results point to important shifts in Flemish cultural journalism. A first clear evolution is that, both in elite and popular newspapers, popular culture topics, media-linked cultural forms and lifestyle topics such as popular music, television, celebrity and culinary topics are becoming increasingly important at the cost of more traditional ‘high-brow’ arts (classical music, poetry, opera, theatre,…). A second and parallel evolution is an advancing horizontal differentiation and changing focus within the most covered cultural forms (performance arts, music, literature, film, television). We notice a proliferation of subgenres within cultural forms evolving from more high-brow/elite to more commercial, democratic and rather low-brow cultural subgenres such as stand-up comedy (performing arts), the comic strip genre (literature), romantic comedy (movies),…
STRIJD OM HET KLIMAAT: DE BEELDVORMING OVER KLIMAATVERANDERING TIJDENS DE KLIMAATTOP TE CANCÚN IN VLAAMSE KWALITEITSKRANTEN - PIETER MAESEELE, YVES PEPERMANS, DANIËLLE RAEIJMAEKERS, LAURENS VAN DER STEEN AND STIJN JOYE Aan de hand van een kwalitatieve inhoudsanalyse wordt in dit artikel de berichtgeving over klimaatverandering verschenen in twee Vlaamse kwaliteitskranten naar aanleiding van de klimaattop te Cancún geanalyseerd en vergeleken. Uit de resultaten blijken belangrijke overeenkomsten en verschillen met betrekking tot de veronderstellingen van waaruit beide kranten vertrekken in hun beeldvorming, in het bijzonder met betrekking tot internationale politiek.
144
145
ETMAAL 2013
BOOK OF ABSTRACTS
AANDACHT VOOR SUÏCIDE IN VLAAMSE EN WAALSE KRANTEN. IMITATIERISICOSCORES EN NALEVING VAN RICHTLIJNEN VOOR JOURNALISTEN - JOLIEN VANGEEL AND ROZANE DE COCK In deze kwantitatieve inhoudsanalyse van acht Vlaamse en Waalse kranten wordt nagegaan in welke mate berichtgeving over suïcide rekening houdt met mediarichtlijnen over dit thema om de kans op imitatie te beperken. Wereldwijd vormt suïcide de 13de meest voorkomende doodsoorzaak met een gestandaardiseerd sterftecijfer van 14,5 per 100.000 inwoners. In vergelijking met andere landen zoals Duitsland, het Verenigd Koninkrijk en Nederland die onder het Europese gemiddelde scoren, kampt België met een veel hoger suïcidecijfer (VAS, 2011, p. 19; VAZG, 2012). In 2008 – het jaar met de meest recente cijfers voor het hele land- maakten 2000 Belgen een einde aan hun leven. Dat aantal ligt ver boven de dodentol die het verkeer in datzelfde jaar eiste (944) (NIS, 2008). Met 5,5 suïcideplegers per dag in België en tot twintig keer zoveel pogingen staat dit maatschappelijk probleem hoog op de agenda van overheid en hulpverleners. In december 2011 werd op initiatief van de minister van Welzijn en Volksgezondheid een tweede Vlaams Actieplan voor Suïcidepreventie (2012-2020) opgezet. In dat plan focust één van de centrale strategieën op “het uitlokken van zelfdoding tegengaan” (VAS, 2012, p 76). Onderzoek wijst immers op de gevaren van blootstelling aan voorbeelden van suïcidaal gedrag in de media, zeker als het gaat om bekende personen (Jamieson, 2002, Jamieson et.al., 2003; Michel et.al; 2000; Motto, 1970; Portsky et al, 2009). Binnen deze beleidsdoelstelling werken door het werkveld en wetenschappers uitgeschreven mediarichtlijnen als indirecte preventie (VAS, 2012, p 35). Deze richtlijnen voor journalisten werden in 2004 voor het eerst opgesteld. Nadat een studie uitwees dat er na de komst van de richtlijnen nog onwetendheid bleef bij journalisten en zij de richtlijnen zelden correct toepasten (Swinnen, 2006), werd de brochure herwerkt (“Als journalist kun je levens redden”, 2007). De richtlijnen kregen ruimere verspreiding onder journalisten en journalistieke opleidingen (VAS, 2012, p 122-123). Een Franstalige versie werd verspreid op initiatief van de Commission de Préventention du Suicide de la Province de Liège. Methode, codeboek en steekproef In deze kwantitatieve inhoudsanalyse gaan we na in welke mate de inspanningen van eind 2007 op vlak van verspreiding van de mediarichtlijnen zich reflecteren in de berichtgeving over zelfdoding in Vlaamse en Waalse kranten (januari –december 2008). Voor Vlaanderen en Wallonië werden telkens twee kwaliteitskranten (De Morgen en De Standaard versus Le Soir en La Libre Belgique) en twee populaire kranten (Het Laatste Nieuws en Het Nieuwsblad versus La Meuse en La Dernière Heure) geselecteerd. Via de persdatabanken Mediargus en Pressbanking werd een overzicht van alle artikelen over suïcide (per krant) gegenereerd. De eerste onderzoeksvraag beantwoorden we aan de hand van het totale aantal artikelen over het topic (N=4303). Voor de andere onderzoeksvragen werden op toevalsbasis 25 artikels per krant geselecteerd, wat een totale steekproef van 200 artikelen oplevert die op basis van de papieren versie werd gecodeerd. Dubbelcodering van 10% van het materiaal door een tweede codeur leverde zeer goede resultaten op voor de intercodeurbetrouwbaarheid (alfa’s tussen .92 en .99). Het codeboek is gebaseerd op een meetinstrument ontwikkeld naar aanleiding van de invoering van mediarichtlijnen over suïcideberichtgeving in Zwitserland (Michel et al., 2000). Swinnen (2006) 145
146
ETMAAL 2013
BOOK OF ABSTRACTS
vertaalde dit voor het eerst naar het Nederlands, zelf voegden we enkele aanvullingen toe om het codeboek te optimaliseren. Het codeboek bevat 63 variabelen waaronder 22 variabelen nodig voor de berekening van de imitatierisicoscore (IRS). Elk van de kenmerken krijgt een score van 0 bij afwezigheid en 1 bij aanwezigheid. De som van deze scores vormt de IRS. Een artikel met een score hoger dan 6 houdt een risico op imitatie in. Onderzoeksvragen OV 1 Hoeveel aandacht besteden Belgische kranten aan suïcideberichtgeving ? OV 2 Op welke manier berichten Belgische kranten over suïcide en is dit in overeenstemming met de mediarichtlijnen ? OV 3 Hoe groot is het aandeel Belgische krantenberichten dat een imitatierisico inhoudt ? OV 4 Hoe groot is het aandeel Belgische krantenberichten dat informatie over preventie en hulpverlening bevat? Bij elke onderzoeksvraag gaan we bovendien na of er significante verschillen zijn in de aanpak tussen Vlaamse en Waalse kranten. Belangrijkste resultaten Onze resultaten tonen aan dat één op drie nieuwsberichten een te hoge imitatierisicoscore heeft en dat slechts één op vijf artikels aandacht besteedt aan hulpverlening en preventie van dit maatschappelijk probleem. Tegen het expliciete advies van de mediarichtlijn in, wordt in 92% van de Vlaamse artikelen toch de term zelfmoord gebruikt en in 75% vermeldt de journalist de methode. Ook privacybeschermende maatregelen worden met de voeten getreden. Zo staat in meer dan de helft van de berichten over specifieke cases de volledige naam van het slachtoffer vermeld.
VLAAMSE MEDIAVERSLAGGEVING OVER CYBERPESTEN - ANNE VERMEULEN AND HEIDI VANDEBOSCH Het onderzoek naar cyberpesten is ondertussen de kinderschoenen ontgroeid. Talrijke kwantitatieve en kwalitatieve studies hebben het probleem al in kaart gebracht (zie bv. Patchin & Hinduja, 2012; Tokunaga, 2010). Momenteel wordt er vooral gekeken naar hoe cyberpesten het best kan worden aangepakt. Daarbij wordt meestal gewezen op het belang van een multi-stakeholder approach, waarin een rol wordt voorzien voor jongeren, scholen, ouders, de politie, ISPs …én de nieuwsmedia (bv. Cross, Li, Smith, & Monks, 2012; Hinduja & Patchin, 2009; Shariff, 2009). Nieuwsmedia kunnen een belangrijke rol spelen in de bewustmaking rond cyberpesten. De hoeveelheid aandacht die zij aan dit onderwerp besteden, kan een invloed hebben op hoe belangrijk het algemene publiek en beleidsmakers dit probleem vinden (cfr. het agendasetting effect) (McCombs & Shaw, 1972). Daarnaast kan ook de manier waarop de media cyberpesten voorstellen of “framen” (door informatie te selecteren, bepaalde oorzaken, effecten en oplossingen te beschrijven,….) de kennis, de attitudes en de acties van de ontvangers sturen (de Vreese, 2005; Scheufele & Tewksbury, 2007). Onder invloed van de media, kunnen verontruste ouders bijvoorbeeld het internetgebruik van hun kinderen meer gaan controleren (Haddon & Stald, 2009). Totnogtoe is er zeer weinig geweten over de mediaverslaggeving over cyberpesten. Enkele vroegere studies tonen aan dat sinds de intrede van het internet, internet-gerelateerde fenomenen een 146
147
ETMAAL 2013
BOOK OF ABSTRACTS
courant thema zijn geworden in de nieuwsmedia. Hoewel sociale netwerksites en het internet vaak in een overwegend positief daglicht worden geplaatst (Rössler, 2001; Vincze, 2011), lijkt bij verslaggeving over het internet en kinderen en/of jongeren het negatieve de boventoon te halen (Haddon & Stald, 2009). Vooral risico’s met betrekking tot sexualiteit en agressie staan hoog op de media-agenda (Haddon & Stald, 2009; Mascheroni, Ponte, Garmendia, Garitaonandia, & Murru, 2010). Voor sommige van deze seksuele en agressie-gerelateerde risico’s geldt dat jongeren vooral het slachtoffer zijn (bv. bij grooming), in andere gevallen zijn ze ook dader (bv. bij cyberpesten en sexting) (Mascheroni et al., 2010). Voorts zijn er enkele indicaties dat de media (gedeeltelijk) gebruik maken van moral panic framing bij de verslaggeving over sexting (Lynn, 2010) en cyberpesten (Mascheroni et al., 2010). In deze paper wordt d.m.v. een exploratieve kwantitatieve inhoudsanalyse van zes Vlaamse dagbladen (twee kwaliteits- en vier populaire kranten) nagegaan in welke mate en hoe de Vlaamse kranten rapporteren over cyberpesten. Gebruikmakend van de online databank Mediargus en verschillende zoektermen, werden alle berichten waarin iets over cyberpesten stond en die waren gepubliceerd voor 1/01/2012, geselecteerd. Na screening op relevantie, resteerden er 449 berichten over (een vorm van) cyberpesten, die vervolgens werden gecodeerd voor een groot aantal indicatoren (datum van publicatie, inhoudelijke klemtoon, geografische focus, betrokken actoren, …). Deze inhoudsanalyse toont dat het allereerste bericht rond cyberpesten reeds werd gepubliceerd in 1998. Cyberpesten is echter vooral sinds 2005, toen de eerste grootschalige studie naar cyberpesten (gefinancierd door de Vlaamse overheid) werd uitgevoerd, een terugkerend onderwerp in de Vlaamse dagbladen. Het krijgt steeds veel aandacht in de maanden september en oktober, wanneer het nieuwe schooljaar start, en in de maanden februari en maart, wanneer de “Vlaamse week tegen pesten” en “Safer Internet Day” plaatsvinden. 182 berichten hadden cyberpesten als hoofdonderwerp en focusten zich voornamelijk op acties, interventies en beleid rond cyberpesten (33%), onderzoek (23,6%) of op één specifieke case van cyberpesten (15,9%). Bij 50 artikelen (27,5%) was de invalshoek geen van bovenstaande. Meestal betrof het dan artikels met een zeer algemene invalshoek waarin werd uitgelegd wat cyberpesten was en waarin bijvoorbeeld enkele cases werden aangehaald. De berichten hadden meestal een nationale (Vlaamse/ Belgische) (55,5%) of een regionale/ lokale focus (30,8%). Slechts 13,7% van de artikelen richtte zich voornamelijk op één of meerdere vreemde landen. Opvallend in de nieuwsberichtgeving rond “cases”, was het grote aantal artikels (7 van de 29) waarin een link tussen slachtofferschap van cyberpesten en zelfmoord werd gelegd. Zoals uit het bovenstaande blijkt, wordt de hoeveelheid en de aard van de berichtgeving over cyberpesten beïnvloed door diverse journalistieke routines. Media hanteren een nieuwsagenda waardoor het nieuws gedeeltelijk voorspelbaar wordt (bv. start van het schooljaar, jaarlijks terugkerende acties …), verkrijgen informatie van diverse bronnen (o.a. onderzoekers, organisatoren van acties) en selecteren (geplande en ongeplande) gebeurtenissen o.b.v. (o.a.) de nieuwswaarden “nabijheid”, “negativiteit” en “het onmiddellijke”. Toekomstig onderzoek kan, vertrekkend van deze eerste inzichten omtrent de mediaverslaggeving 147
148
ETMAAL 2013
BOOK OF ABSTRACTS
rond cyberpesten, nagaan welke invloed deze heeft op de kennis, de attitudes, de risicoperceptie, en de uiteindelijke gedragingen van bv. jongeren en hun ouders.
Session #9 Friday 8 February 11:45 – 12:45 hrs. Chair: Erik Hitters Location: T3-29 Dutch session
DIVERSITEIT IN HET VLAAMSE PERSLANDSCHAP: DOMINEREN BELGISCHE VOLWASSEN MANNEN ALS ELITEBRONNEN IN DE HARDERE NIEUWSTHEMA’S? - HANNE VANDENBERGHE, LEEN D'HAENENS AND BALDWIN VAN GORP Situering In deze paper wordt open diversiteit in de Vlaamse pers op het vlak van actoren en thema’s onder de loep genomen. Deze studie focust zich op de reflectionistische kijk op mediadiversiteit, in tegenstelling tot de constructivistische stroming die elk mediaproduct als een construct op zichzelf, los van de werkelijkheid, ziet (Orgad, 2012). Binnen die eerste stroming stelt van Cuilenburg (2005) dat massamedia verantwoordelijk zijn om alle heersende opvattingen binnen een maatschappij op een evenredige wijze in media-inhoud te representeren. Vertaald naar nieuws betekent dit dat nieuwsbronnen demografisch conform de reële samenleving dienen verdeeld te zijn. Literatuur De ondervertegenwoordiging van vrouwen in nieuwsinhoud is in tal van onderzoek vastgesteld, zowel in het Vlaamse televisienieuws (De Swert & Hooghe, 2010) als in een Amerikaanse krantencontext (Len-Ríos, Rodgers, Thorson, & Yoon, 2005). Daarnaast blijken etnische minderheden in Europese en Nederlandse berichtgeving weinig over alle thema’s heen voor te komen, maar wel iets hogere percentages te halen in negatief nieuws of criminaliteitsnieuws (ter Wal, d’Haenens & Koeman, 2005). Wat leeftijd betreft, blijkt uit een Portugees-Europees onderzoek (Ponte, 2007) op basis van de mainstream pers dat kinderen vaak als weerloze slachtoffers voorkomen en zelf geen uitspraken doen. Oudere vrouwen in de Argentijnse massamedia komen daarnaast weinig voor en worden vaak stereotiep negatief of veel jonger afgebeeld (Teubal, 2000). Onderzoeksvragen en verwachtingen OV1: Wat zijn de verhoudingen binnen de opgevoerde actoren qua leeftijd in de Vlaamse dagbladpers? OV2: Wat is de man-vrouwverdeling van opgevoerde actoren in de Vlaamse dagbladpers? OV3: Wat is de verhouding tussen etnische minder- en meerderheden in de opgevoerde actoren in de Vlaamse dagbladpers? OV4: Welke functie-indeling (elitebron, onvervangbare non-elitebron/getuige of vervangbare nonelitebron/ voxpop) vervullen deze ‘groepen’ actoren in de Vlaamse pers? Hierbij is een elitebron onvervangbaar omwille van de beroepssituatie, geeft een onvervangbare non-elitebron relevante of 148
149
ETMAAL 2013
BOOK OF ABSTRACTS
exclusieve informatie vrij en is een vervangbare non-elitebron inwisselbaar voor een ander individu. OV5: Worden er specifieke thema’s geassocieerd met leeftijdsgroepen (jongeren, volwassenen en ouderen), mannen of vrouwen en etnische minder- of meerderheden? De focus ligt telkens op binnenlands nieuws (d.w.z. in een Belgische context). Op die manier kan de demografische realiteit met de voorstelling in de pers worden vergeleken. Methodologie Via inhoudsanalyse met een a priori vastgelegd codeerschema (gebaseerd op o.a. Koeman, Peeters, & d’Haenens, 2007 en Devroe, 2007) werden alle krantenartikelen van een geconstrueerde lenteweek (tussen 3 april en 19 mei 2012) bij alle Vlaamse dagbladen (drie kwaliteits-, twee publieks, één gratis en twee regionale kranten) geanalyseerd (N = ca. 6000). Columns, editorialen, lezersbrieven en recensies werden weggelaten. Alle individuen worden ingedeeld op basis van leeftijd, gender en etniciteit. Voor leeftijd zijn drie leeftijdscategorieën – jongeren staan voor min 18-jarigen, ouderen voor de 65-plussers en volwassenen voor de tussencategorie – vooropgesteld. Indeling gebeurt op basis van beroepsstatus of beschrijving. Bij vermelding van de exacte leeftijd in het artikel wordt dit ook opgenomen. Daarnaast werd het geslacht gecodeerd indien de journalist dit letterlijk vermeldt, of, waar mogelijk, afgeleid uit de naam of functie. Etniciteit tot slot is samengesteld uit vermelde nationaliteit en een open variabele waarin alle herkomst gerelateerde uitdrukkingen worden opgesomd. Voorbeelden van dergelijke associaties zijn het vermelden van termen als ‘racisme’, ‘allochtoon’ of ‘illegaal’, een weergave van een andere religie of een verwijzing naar integratie. Resultaten Uiteindelijke resultaten worden momenteel nog gegenereerd. De voorlopige resultaten, op basis van een eerste, kleine steekproef van lange artikelen (d.w.z. meer dan 200 woorden) van dinsdag 3 april 2012 van de kwaliteitskrant De Morgen leverde 125 individuen in een Belgische context op met onderstaande diversiteitsverdelingen. Qua leeftijd werden 2 jongeren (1,6%), 116 volwassenen (92,8%) en 2 ouderen (1,6%) geteld. Voor vijf personen kon geen leeftijdscategorie worden bepaald. In verhouding met de realiteit blijken jongeren en ouderen sterk ondervertegenwoordigd: de werkelijke verhouding in 2011 was namelijk 19,5% jongeren, 62,2% volwassenen en 18,3% ouderen (FOD Economie, 2011). De twee jongeren zijn getuigen in sportnieuws en in nieuws over sociale zaken. De twee ouderen kwamen voor in het cultuurthema als elitebronnen. De genderverdeling kwam uit op 101 mannen (80,8%) t.o.v. 23 vrouwen (18,4%) – bij één actor kon geen gender worden vastgelegd – in tegenstelling tot de reële sleutel van 49,5% mannen versus 50,6% vrouwen (FOD Economie, 2011). Hierbij waren 18 vrouwen opgevoerd als beroepsmatige elitebron t.o.v. 96 mannen. Qua nationaliteit wordt voor 10 individuen een niet-Belgische nationaliteit vrijgegeven, waarvan er vier West-Europees zijn. Er werden geen herkomst gerelateerde elementen opgegeven. 149
150
ETMAAL 2013
BOOK OF ABSTRACTS
Zowel binnen een economische en culturele context kwamen drie ‘vreemde’ individuen voor, bij sport en toerisme respectievelijk twee. Allemaal werden ze als een elitebron gepresenteerd. Besluit Voorlopig betreft het hier een zeer beperkte steekproef als voorproef op de resultaten die in februari 2013 worden gepresenteerd. Daarom doen we nog geen uitspraken over thema’s of functie-indeling. Enigszins voorbarig stellen we wel dat Belgische volwassen mannen de Vlaamse pers domineren. Dit impliceert dat het nieuws de diverse realiteit disproportioneel afspiegelt.
ACTOR DIVERSITY IN FLEMISH ECONOMIC TELEVISION NEWS - ANOESKA SCHIPPER, HILDE VAN DEN BULCK AND STEFAAN WALGRAVE With the 2008 financial and subsequent euro crisis, economic news seems to have gained prominence. At the same time, the complexity of the issues concerned raise questions about the quality of the coverage of economic issues. The paper deals with one aspect of the quality of economic news coverage. It assesses the diversity of the actors in economic news and accounts for the varying levels of actor diversity in economic news items broadcasted between 2006 and 2011 on the public and commercial channels in Belgium (Flanders). News actors and journalists both have a stake in the news. News actors need a platform to express their views and journalists need actors that are available, suitable, and reliable. The problem is that these mutual interests can lead to journalistic routines favoring the presence of certain actors in the economic news and resulting in an imbalance of news actors and a decrease of the diversity of news actors. We argue that the type of media organization, a journalist’s expertise, the specific type of economic issue and the economic period are determinants for the diversity of news actors in the economic news coverage. First, on a general level, we expect there to be a difference between commercial and public media organizations having different budgets leading to differences in populated newsrooms, time pressure and possibly a different selection of news actors. Second, we expect that the expertise of journalists in economic news is another predictor, as journalists with a considerable expertise in economic news may have closer ties with specific economic news actors and may use less diverse news actors to get their story across. Third, we assume that the link that news actors have with the type of issue, such as the link between financial news and investors, banks and rating agencies determines the diversity of news actors. Last, we hypothesize that in economic periods of stability there is more room for diverse news actors than in periods of economic crises and uncertainty, when journalists may depend more on reliable news actors with power and status. Diversity means that at least two news actors with different roles are directly quoted. This is the dependent variable. The independent variables are 1) the type of media organization (commercial or public), 2) the degree of expertise of journalists in economic news coverage, 3) the type of issue (financial, ‘pure’ economic, employment and consumer affairs), and 4) the different periods within the covered period (economic prosperity before the crisis, the financial crisis, and the subsequent euro crisis).
150
151
ETMAAL 2013
BOOK OF ABSTRACTS
The empirical analyses are based on a quantitative content analysis of all prime-time (19h00) news casts covering economic news on the two main Belgian news stations. The sample for this analysis is taken from the database belonging to the Flanders Media Policy Research Centre, that archives and encodes television news in Flanders. Preliminary results suggest that journalists with expertise in economic news use not more diverse news actors than journalists with little expertise. The type of media organization and the economic theme do not result in more actor diversity. Results indicate that more diverse news actors are selected during the credit crisis and the euro crisis than in the period of economic prosperity.
'DE DOCUMENTAIRE IS DOOD, LANG LEVE DE DOCUMENTAIRE.' DE VLAAMSE ONAFHANKELIJKE DOCUMENTAIRESECTOR OP ZOEK NAAR EEN NIEUW ELAN? - DORUNTINA ISLAMAJ België kan terugblikken op een lange documentairetraditie, denken we maar aan de sociaal bewogen documentaires uit de jaren ’30 van Henri Storck, Charles Dekeukeleire en Joris Evens. Een documentaireschool die omschreven werd door de bekende documentairemaker Robert Flaherty als ‘de interessantste van de wereld’ (Robert Flaherty, 1938). Desalniettemin kunnen we vaststellen dat het sinds de jaren ‘80 moeilijk is om het genre aan de man te brengen. Het eens gevrijwaarde klimaat kwam op de helling te staan: marktlogica’s en kijkcijfers haalden samen met privatiseringen de bovenhand. De toename van het aantal zenders zorgde er eveneens voor dat de concurrentie toenam en dit dwong zenders andere koersen te bevaren. Nieuwe programmastrategieën werden ontwikkeld waarbij de rol van documentaire vaak ter discussie stond. Het bedoelde maatschappelijk debat kreeg minder en minder een plaats en leek mettertijd niet meer dan een nichepubliek aan te spreken. Vanaf de jaren ‘90 was het documentairegenre dan ook op de achtergrond terecht gekomen. Op vele plaatsen kende men geen goed werkende filmfondsen, omroepen voerden een strijd om kijkcijfers en door de liberaliseringen in de mediasector met de daarbij horende stroom aan veranderingen moest het genre zich noodzakelijk gaan herpositioneren. Documentaire werd hierbij nog al te vaak gepercipieerd als een plaatsgebonden en commercieel inert genre, als een anomalie in de filmmarktplaats. Vertrekkend vanuit dit historisch perspectief wil deze paper de Vlaamse onafhankelijke documentaireproductie in kaart te brengen en onderzoeken welke kansen deze sector vandaag krijgt binnen een Vlaamse en Europese beleidscontext. De problematiek wordt empirisch benaderd met een politiek-economische inslag. Tussen de periode 2010-2012 werden in totaal 17 diepte-interviews en tal van gesprekken gevoerd met verschillende actoren uit de sector. Samen met een grondige studie van de verschillende beleidsdocumenten werd er op zoek gegaan naar de knelpunten die zich vandaag in de Vlaamse onafhankelijke documentairesector aandienen op zowel productie-, distributie- en exploitatie gebied. Met als hoofdvraag: welke toekomst heeft de Vlaamse documentaire anno 2012? De resultaten geven aan dat sinds 2000 de gezondheidstoestand van de Vlaamse filmsector, voornamelijk door de oprichting van het Vlaams Audiovisueel Fonds, er fors op vooruit gegaan is. De documentairesector blijft tal van moeilijkheden kennen, maar er zijn ook diverse specifieke steunmechanismen operatief en het lijkt erop dat een nieuwe generatie Vlaamse documentairemakers is opgestaan. Vooral de oprichting van het Vlaams Audiovisueel Fonds (VAF), de 151
152
ETMAAL 2013
BOOK OF ABSTRACTS
invoering van de tax shelter en de toegenomen internationale mogelijkheden betekenden hierin de belangrijkste stimulansen. Terzelfdertijd wordt de rol van de omroep als stimulator en afnemer van documentaires problematischer. Daar zit dan ook meteen de grootste uitdaging voor de Vlaamse onafhankelijke documentaire vandaag, het vinden van mogelijke distributiekanalen. Vlaamse documentaires vinden bijzonder moeilijk afzet binnen de wetmatigheden van televisie en dringen ook nauwelijks door naar de cinemazalen. Een trend waardoor Vlaamse documentaires voornamelijk op festivals getoond worden en slechts zelden terugvloeien naar het grote publiek. Wanneer gekeken wordt naar de beperkte distributiemogelijkheden is de belangrijkste vraag die zich opwerpt of de toekomst van documentaire in een ‘nichepubliek’ zal zitten of ‘een breder publiek’ kan aanspreken. Ook de rol van internetdistributie kan hierin een mogelijk piste worden. Ondanks de diverse problemen waar onafhankelijke makers mee geconfronteerd worden blijkt weliswaar uit de interviews dat er een nieuwe generatie Vlaamse documentairemakers is opgestaan die verenigd in ‘Flanders doc’ het genre een nieuw gezicht willen geven en terug sterk wil laten staan op de internationale markt. Onze Vlaamse creatieve documentaires – die nog steeds begeesterd en vindingrijk worden gemaakt – hebben toekomst, maar die veiligstellen zal niet zonder slag of stoot gaan. Een soepelere verspreiding en enthousiasmerende nieuwe manieren van vertoning, daarin kunnen de nieuwe sleutels liggen tot een nieuw hoofdstuk in de documentairetraditie van België.
152
153
ETMAAL 2013
BOOK OF ABSTRACTS
ORGANIZATIONAL COMMUNICATION Session #1 Thursday 7 February 14:05 – 15:05 hrs. Chair: Ericka Menchen-Trevino Location: T3-34 English session
DEALING WITH DIFFERENCES: THE IMPACT OF DIVERSITY PERCEPTIONS ON ASSESSMENT OF MINORITY EMPLOYEES - JOEP HOFHUIS, KAREN VAN DER ZEE AND SABINE OTTEN An important issue in organizational research, which has steadily gained attention in the past decades, is the promotion and management of cultural diversity in the workplace. Scholars have emphasized that the presence of employees with different cultural backgrounds on the work floor may be vital to an organization’s competitiveness. Through increased knowledge sharing and idea generation, diverse employees may enhance innovation and adaptability of their work teams. Although many strategies have been employed to specifically recruit and select minority employees, the selection rates for designated minority groups are often lower than those for the majority group. Particularly minority candidates with high cultural maintenance (those who strongly identify with their cultural and/or ethnic background, and choose to openly display and express their heritage) are vulnerable to cultural bias in selection procedures. In this paper, research is presented which shows that this bias may be circumvented by promoting a positive perception of diversity among recruiters. Those who are aware of the possible benefits of diversity may be more open to selecting minority candidates than those who view diversity as threatening to the organization. In an experimental study, participants (n = 99) belonging to the cultural majority group played the role of recruiters. Their diversity perceptions were manipulated by asking them to think about, and discuss in groups, either positive or negative outcomes of cultural diversity in the workplace. They were then asked to rate minority candidates for a fictional job opening. The candidates’ profiles varied in the degree of cultural maintenance. Firstly, our results confirmed that cultural maintenance of minority candidates has a negative effect on the ratings they receive in assessment procedures. Overall, recruiters are less positive towards candidates who choose to retain their own culture in the workplace. More importantly, this effect is moderated by diversity perceptions. Recruiters who perceive individual differences in the workplace as positive and beneficial, give higher ratings to candidates who maintain their own culture than those who perceive diversity as threatening. Our research shows that recruiters who believe there is value in diversity may be more successful at spotting potential talent among minority candidates. This implies that organizations aiming to take advantage of diversity in the workplace, should make an effort to communicate the positive outcomes of cultural differences to employees involved in selection and assessment. The results may be twofold: it should increase the selection of employees with a different cultural background, as 153
154
ETMAAL 2013
BOOK OF ABSTRACTS
well as prepare the organization for actually benefiting from this increased diversity in terms of innovation and competitiveness.
ENGAGING EMPLOYEES IN CORPORATE CITIZENSHIP ACTIVITIES: THE CASE OF IBM INDIA - VIDHI CHAUDHRI AND ASHA KAUL Conversations about corporate social responsibility (CSR)/ corporate citizenship (CC) in organizations have spurred an interest in employee engagement in CSR as a mechanism for higher employee identification, satisfaction, and retention (e.g., Greening & Turban, 2000). Citing employee engagement in CSR/CC as a strategic imperative and a source of competitive advantage, many organizations now allow employees time off from work to volunteer in the community (e.g., Smith & Sypher, 2010). Indeed, Bhattacharya, Sen, and Korschun (2008) suggest that CSR can be a “genuine point of differentiation” for a company and can be used to enhance the “employee value proposition” by creating the feeling that they are part of a larger corporate mission (p. 37). CSRbased identification, they note, generates numerous benefits for the organization including higher employee morale, commitment, satisfaction, recruitment, retention, and a reduction in absenteeism. Employees, too, are seeking opportunities that allow them to make a difference through their jobs (Net Impact, 2012). A survey by Net Impact (2012) found that respondents who have ‘impact’ opportunities at work (“or jobs that provide the opportunity to make social or environmental impact”) are likely to be twice as satisfied as those whose work lacks impact opportunities (p. 2).1 Despite the purported benefits of employee engagement, there is limited understanding of how organizations create opportunities to involve employees in CSR/CC that go beyond sporadic and adhoc volunteerism. Thus, our study examines how organizations create long-term and sustainable strategies of engagement that are embedded in organizational culture. In doing so, how do organizations articulate and communicate shared and collective beliefs and values of social responsibility? A related goal is to understand the (perceived) benefits—for organizations and their members—of fostering an organizational culture of social responsibility. We present an in-depth case study of IBM India’s approach to employee engagement. Globally, IBM has been recognized as a CC leader with societal initiatives in the areas of education, healthcare, and environment that leverage and align with the organization’s technological strengths and innovation capabilities. Former CEO Sam Palmisano asserted that social responsibility at IBM is “a distinct management approach, encompassing investment, talent, policy, governance and stakeholder engagement” that is embedded in the cultural fabric of the organization, and manifested in the everyday actions of ‘IBMers.’2 Using a combination of company data, executive interviews, and focus group discussion with a group of employees at IBM India, the case illuminates the multitude of factors –structural, cognitive, behavioral, and relational—which have facilitated the organization’s ability to build social capital. We find that by integrating its business and social strategies and by fostering a culture of responsibility, 1
Net Impact. (2012). Talent Report: What Workers Want in 2012. Retrieved October 19, 2012 from http://netimpact.org/docs/publications-docs/NetImpact_WhatWorkersWant2012.pdf. 2 A Letter from Samuel J. Palmisano.Retrieved June 26, 2012 from http://www.ibm.com/ibm/responsibility/report/2010/chairmans-letter/index.html. 154
155
ETMAAL 2013
BOOK OF ABSTRACTS
IBM India is able to position itself as a ‘responsible leader,’ attract and retain talent, derive brand value, and promote opportunities for collaboration and shared learning across the global organization. The study also finds a central role for communication in fostering a culture of corporate responsibility and a strong sense of association with an organization that does more than make profits. Especially organizational narratives and storytelling emerge as the foremost mechanism for socializing new volunteers, inviting participation, and showcasing lessons learned and (personal and professional) benefits of the organization’s culture of service. In sum, the case provides an in-depth look at how a global organization translates its CC vision and strategy across the countries where it operates. In so doing, the case offer insights—from an organizational communication and culture perspective—that may be transferable to other organizations seeking active employee engagement.
GREAT EXPECTATIONS: THE ROLE OF MENTORING SCHEMAS IN PROTÉGÉS' AND MENTORS' SENSEMAKING OF FORMAL MENTORING RELATIONSHIPS - SUZANNE JANSSEN, MARLEEN MENSINK AND MARK VAN VUUREN Introduction Formal mentoring programs are associated with positive effects for protégés, mentors, and organizations. While a good match of the role expectancies of both the mentor and the protégé is crucial for the effectiveness of the relationship, research on this matching process is scarce. Current research has mainly focused on dispositional and demographic variables, while ignoring relational factors. Based on Baldwin’s framework of relational schemas (Baldwin, 1992) and relational identity and identification (Sluss & Ashforth, 2007), this study applied a relational lens to examine how mentoring schemas influence participants’ perceptions and interactions within formal mentoring relationships. Mentoring schemas were defined as “cognitive maps derived from past experiences and relationships that guide mentor’s and protégé’s perceptions, expectations, and behaviors in mentoring relationships” (Ragins & Verbos, 2007, p. 101). These mentoring schemas include generic mental representations about the general roles of mentors and protégés (i.e., “Mentors are experts”) as well as typical mental representations reflecting mentor and protégé roles in their specific relationship (i.e., “The protégé takes the lead during conversations”). The aim of this study was to gain understanding in the ways such schemas influence members’ sensemaking processes of the mentoring relationship. Consequently, our research questions were: (a) How do protégés and mentors perceive the role expectations from themselves and the other at the beginning of their relationship? (b) How do these expectations influence the initial evaluation of the relationship? (c) How do these mutual expectations and schemes evolve over time? And (d) What is in retrospect seen as the contribution of these schemas to the impact and quality of the relationship? Method Mentors and protégés (n=22) from a formal mentoring program at a Dutch university were interviewed about (a) their perceptions of what constitutes a good mentor, a good protégé, and a 155
156
ETMAAL 2013
BOOK OF ABSTRACTS
good mentoring relationship, (b) their initial expectations from their partner, themselves, and their relationship, (c) the behaviors typically exhibited in the relationship, and (d) their evaluations of the behaviors and quality of the relationship. Results First, we sketch out the mental representations that are held by mentors and protégés. While both mentors and protégés are focused at career development for the protégé as primary goal of the relationship, their expectations of behaviors exhibited in the relationship differ. Both mentors and protégés expect the other to take the lead. While mentors and protégés reported clear expectations of their partner, expectations from themselves in the relationship were often unclear. While mentors describe the relationship in terms of communal norms and therefore do not expect anything in return from the protégé, protégés apply an instrumental approach to the relationship and expect the mentor’s efforts to be beneficial for the protégé. Second, based on reported initial expectations, three types of mentoring schemas were identified: fixed, flexible, and open. Members with fixed schemas reported clear role expectations of both their partner and the expected outcome of the relationship. Members with flexible schemas did have some expectations, but were willing to adapt those to their partner, depending on the situation. Members with open schemas entered without expectations, and decided to “cross that bridge when we get there.” Third, results show narratives of how mentoring schemas influenced members’ behaviors in the relationship, their satisfaction with the relationship, and their overall evaluation of the relationship’s quality and effectiveness. This was partly due to the initial schemas. For example, while protégés with a fixed initial schema reported being satisfied with the relationship, protégés with open schemas reported being unsatisfied. Moreover, results show how mentors and protégés who have congruent mentoring schemas perceive the relationship as more satisfactory than members whose initial relational schemas did not match (e.g. both considered themselves as reactive party). Conclusion This study shows which relational schemas are held by members of a formal mentoring program and how these schemas influence sensemaking processes of relational roles and relational functioning.
AUDIENCE BUILDING BY POLICE OFFICERS ON TWITTER: DO FOLLOWERS FOLLOW FOR BUSINESS OR PLEASURE? - BOB VAN DE VELDE AND IVAR VERMEULEN Police use of twitter hinges on reaching a big audience. We use uses and gratification theory to examine how user and message characteristics, such as activity and topic, help build an audience. We examine this process with observed twitter use of 964 accounts and their messages. A linear regression predicting audience size shows informative characteristics have a positive effect, whereas small talk has a negative effect.
156
157
ETMAAL 2013
BOOK OF ABSTRACTS
Session #2 Thursday 7 February 15:15 – 16:15 hrs. Chair: Marc Verboord Location: T3-34 Dutch session
MEDEWERKERS ALS MERKAMBASSADEURS - JOOST VERHOEVEN AND PIET VERHOEVEN Door de opkomst van sociale media heeft het internet steeds meer de massamedia vervangen als de arena waarin merken worden gemaakt en gekraakt (Chaffey, at al., 2009). Op sociale media kunnen gebruikers zonder tussenkomst van een redactie een merkgerelateerde boodschap ontwerpen en verspreiden. Dit maakt het voor organisaties steeds moeilijker om merkcommunicatie te controleren. In marketing wordt de aandacht daarom verlegd van massamediale campagnes naar het creëren van enthousiaste, loyale fans die als merkambassadeur het organisatiemerk te versterken. Daarbij wordt de rol van medewerkers van de organisatie nog vaak over het hoofd gezien. Toch zijn medewerkers erg actief op sociale media en vertegenwoordigen zij op dergelijke platforms de organisatie. Communicatiemedewerkers gebruiken bijvoorbeeld corporate Facebookpagina’s, Twitteraccounts en LinkedIngroepen voor het verspreiden van (marketing-)boodschappen en het voeren van een dialoog met gebruikers. Echter, niet alleen communicatiemedewerkers, maar alle leden van een organisatie kunnen online - al dan niet bewust - bijdragen aan het organisatiemerk. Daarvoor gebruiken medewerkers meestal niet de officiële socialemediakanalen van de organisatie, maar hun persoonlijke socialemedia-accounts. Zij doen dit veelal op eigen initiatief en buiten de controle van de organisatie om. Er is nog weinig inzicht in de mate waarin, en de manier waarop medewerkers het organisatiemerk beïnvloeden door op sociale media over hun werk en over hun werkgever te berichten. Om de omvang en betekenis van deze informatiestroom in kaart te brengen is onder 375 medewerkers van 8 organisaties een online survey afgenomen. Daarnaast is de inhoud van 3718 tweets van medewerkers geanalyseerd. Uit de survey blijkt dat vrijwel alle medewerkers actief zijn op sociale media. Hoewel de meesten sociale media vooral voor privécontacten gebruiken, publiceert ongeveer een derde van de medewerkers berichten die gaan over het werk dat ze doen, de organisatie waar ze werken, of de producten en diensten die de organisatie levert. De inhoudsanalyse heeft aangetoond dat ruim 20% van de geanalyseerde tweets gaat over het werk of de werkgever van de twitteraar. Deze boodschappen gaan vooral over de dagelijkse werkzaamheden die medewerkers vervullen en laten daarmee zien welke issues op dat moment belangrijk zijn voor hen en voor de organisatie. Deze tweets geven de organisatie een menselijk, authentiek gezicht. Ook berichten medewerkers over actuele issues zoals evenementen, productinnovaties, marketingcampagnes of vacatures. Medewerkers geven bekendheid aan deze issues en beïnvloeden daarmee de publieksagenda en de associaties van het publiek met het organisatiemerk. Omdat medewerkers grotendeels autonoom zijn in hun socialemediagedrag kunnen merkmanagers geen directe controle houden over de sociale media berichten. Toch kunnen organisaties deze 157
158
ETMAAL 2013
BOOK OF ABSTRACTS
communicatiestroom in goede banen proberen te leiden door het sociale media gedrag van medewerkers te monitoren en daarvan te leren, door medewerkers te stimuleren om over hun organisatie te praten op sociale media en door duidelijke communicatieregels te hanteren m.b.t. informatieontsluiting. In een veranderde omgeving, waarin transparantie en authenticiteit belangrijke waarden zijn, is het essentieel voor merkmanagers om een bottom-up merkstrategie te hanteren om het organisatiemerk te versterken. Daarbij spelen medewerkers een cruciale rol in het versterken van het online corporate merk. Door hun activiteiten op sociale media te monitoren, hen te stimuleren op een verstandige en verantwoorde manier sociale media te gebruiken en als merkambassadeurs deel te nemen aan de merkdialoog, kunnen organisaties op een geloofwaardige manier aan een sterk organisatiemerk bouwen.
HOE CREËER JE MERKTROUW? EEN CONFRONTATIE TUSSEN VISIES UIT WETENSCHAP EN PRAKTIJK - SUZANNE DE BAKKER AND PEETER W.J. VERLEGH Laten we starten met een open deur: loyale klanten zijn van levensbelang voor organisaties die een product of dienst verkopen. Zo besteden loyale klanten bijvoorbeeld meer dan gewone klanten (Reicheld & Teal, 1996), zijn ze meer geneigd om netwerken van vrienden, familie en andere potentiële klanten aan de organisatie te verbinden (Shoemaker & Lewis, 1999), en zijn ze minder vatbaar voor promotie-acties van concurrerende merken en organisaties (Reicheld & Teal, 1996). Maar hoe creëer je als marketingmanager loyaliteit? Welke communicatie-instrumenten zet je in, en op welke manier worden consumenten door deze instrumenten beïnvloed? In dit onderzoek proberen we een antwoord te krijgen op deze vraag, door literatuurstudie en interviews met managers van 20 bekende Nederlandse merken. Een literatuurstudie laat zien dat er veel verschillende visies bestaan op het concept, maar dat men het model van Oliver (1999) vaak als uitgangspunt neemt. Dit model beschrijft de vorming van loyaliteit in vier opeenvolgende fasen, waarbij elke fase een sterkere vorm van loyaliteit kent. Volgens Oliver is de eerste fase in het loyaliteitsproces de ‘cognitive loyalty’. Hierbij is sprake van (herhaal)aankoopgedrag, maar niet van emotionele binding. De aankoop vindt plaats vanuit een gewoonte, gemak of een andere reden en is vaak gebaseerd op (cognitieve) productinformatie. Als de transactie routinematig is, blijft de tevredenheid onbewust. Loyaliteit is dan meer dan een (gewoonte)gedrag. Als de tevredenheid echter wel bewust is, wordt deze onderdeel van de klantervaring en ontstaat er emotionele betrokkenheid bij het merk. Deze tweede fase noemt Oliver ‘affective loyalty’. Net als in de cognitieve fase, komt hier nog brandswitching voor, en is de consument niet 100% trouw aan het merk. Het switchgedrag wordt al minder in de derde fase, ‘conative loyalty’. Dit is de gedragsintentie-fase, die een merkspecifieke betrokkenheid bij een herhaalaankoop impliceert. In de vierde ‘action loyalty’ fase wordt deze intentie omgezet in actie. Naast deze vier fasen beschrijven Gounaris & Stathakopoulos (2004) ‘covetous loyalty’: loyaliteit met een zeer hoge affectieve betrokkenheid maar zonder(herhaal)aankoopgedrag. Hierbij is het vaak niet de keuze van de consument zelf om het merk of product niet te kopen: denk bijvoorbeeld aan Ferrariliefhebbers met te weinig geld voor een Ferrari.
158
159
ETMAAL 2013
BOOK OF ABSTRACTS
Hoe kun je nu als marketeer, door middel van marketingcommunicatie, dit actieve proces van merkloyaliteit ondersteunen? Tot pakweg vijf jaar geleden was het nog ‘makkelijk’: volgens HennigThurau en collega’s (2010) beïnvloedden merken voorheen de klantrelaties door marketingacties zoals loyaliteitsprogramma’s, maar ook door zowel reactieve als proactieve public relations. Klanten werden gezien als passieve ontvangers van marketing en communicatie en merken hadden bijna volledige controle over hun merkvormende boodschappen. Echter, door de mediaveranderingen van de afgelopen jaren lijkt deze visie verouderd. De vraag die dan centraal staat is: hoe kunnen merken, in deze tijd van sociale media en dialoog, merkloyaliteit ontwikkelen en behouden? Deze vraag is beantwoord in twintig interviews met marketing- en brandmanagers van merken uit de FMCG, banken en verzekeringen, telecomaanbieders en energiesector. Langs de lijnen van het model van Goodall (2009) lag de focus daarbij op de inzet van betaalde media (zoals televisiereclame), eigen media (klantmagazines, maar ook bijvoorbeeld websites) en de rol van verdiende aandacht (bijvoorbeeld via sociale media – zie Weima (2012)). Uit deze interviews komt naar voren dat merkloyaliteit niet zozeer een expliciet geformuleerd doel is, maar dat het wel de basis is van het voortbestaan van het merk. Ondanks de aandacht vanuit de wetenschap aan de rol van emoties en merkliefde denken managers een stuk nuchterder over loyaliteit. Cognitieve loyaliteit wordt gezien als de vorm van loyaliteit die het meest voor de hand ligt, waarbij, als het even kan, affectieve gevoelens ontwikkeld kunnen worden. Opvallend is de nadruk die managers leggen op ‘real life’ ervaringen van klanten. Zij proberen klanten de merkwaarden te laten ervaren door bijvoorbeeld het organiseren en/of sponsoren van evenementen, en het kopen van ruimte in winkels om mensen het merk te laten ervaren. In conclusie kan gesteld worden dat de loyale klant een begerenswaardige consument is, die zich misschien makkelijker (door wetenschappers) laat beschrijven dan (door praktijkmensen) laat vangen. Door wetenschap en praktijk met elkaar te confronteren probeert dit onderzoek een brug te slaan tussen de verschillende visies op merktrouw. Hierbij wordt duidelijk dat het (uitzonderingen daargelaten) moeilijk is om de uitgebreide en diepgaande beschrijvingen van dit concept te vertalen naar de nuchtere werkelijkheid van switchende en weinig gecommitteerde klanten. Het lijkt voor de meeste merken dan ook aan te raden om de pijlen te richten op de meer basale vormen van loyaliteit, en hogere vormen vooral te zien als een inspirerend ideaalbeeld.
HET 7I NETWERKMODEL – 7 KRITISCHE SUCCESFACTOREN VOOR EEN 3.0 ORGANISATIE - RENEE VAN OS, JANNEKE DELISSE, MARCO DERKSEN AND JAN JURRIËNS Inleiding De laatste jaren zijn het gebruik van sociale media en het daaraan gekoppelde samenwerken in (online) netwerken sterk toegenomen. Echter, bij veel organisaties lijkt een onderliggende visie of strategie te ontbreken, met als gevolg: applicaties als Twitter, Facebook en YouTube worden redelijk wille-keurig ingezet, en met wisselend succes. De vraag die leeft in zowel wetenschap als praktijk: passen deze organisaties sociale media wel adequaat toe? En nog belangrijker: zijn organisaties eigenlijk wel klaar voor de inzet van sociale media? Om deze situatie nader te bestuderen is het ‘’7I netwerkmodel’’ ontwikkeld en gevalideerd vanuit 159
160
ETMAAL 2013
BOOK OF ABSTRACTS
zowel de theorie als de praktijk. Het model kent zeven indicatoren die samen de 3.0 organisatie in kaart brengen: een organisatie waarin interacties tussen organisaties en hun stakeholders centraal staan, met inzet van sociale media (Van het Hoff, 2011). Uitgangspunt van het 7I netwerkmodel is dat een organisatie pas succesvol kan zijn in het opbouwen en onderhouden van een netwerk met stakehol-ders via sociale media (en dus een 3.0 organisatie is) wanneer intern (identiteit, structuur en cultuur) en extern (interactie met stakeholders) aan enkele randvoorwaarden is voldaan. Dit in tegenstelling tot veel andere sociale media modellen die uitsluitend ingaan op externe factoren (e.g. De Lange, Hor-dijk, Kaassenbrood, 2010; Kerkhofs, Beerthuyzen en Adriaanse, 2011). Het 7I netwerkmodel is begin 2011 ontwikkeld voor de adviespraktijk. Op dat moment ontbrak een uitgewerkte operationalisatie van de indicatoren. Begin 2012 is de HAN in samenwerking met de ontwikkelaars gestart met het valideren van het model middels een theoretische verkenning en een pilot-onderzoek onder zeven organisaties. Dit paper gaat in op de resultaten van zowel de theoretische verkenning als het pilot-onderzoek. Theoretische verkenning – wat ‘maakt’ een 3.0 organisatie? Het theoretisch fundament voor het 7I netwerkmodel is gevonden in de literatuur over ´netwerkorganisaties´. Onderzoek naar netwerkorganisaties gaat decennia terug. Zo identificeerden Bernassi, Greve en Harkola (1999) de volgende indicatoren van een netwerkorganisatie: platte, flexibele organisatiestructuur, toegankelijk, focus op informatie-uitwisseling en kennisdeling, zelfsturende medewerkers, veelheid aan externe relaties waarmee gecommuniceerd wordt. En reeds in 1979 sprak Mintzberg van een ‘adhocratie’ als organisatievorm waarin horizontale relaties, flexibele structuren en groot onderling vertrouwen en respect centraal staan. De kern van een netwerkorganisatie volgens diverse (meer recente) onderzoeksliteratuur betreft: het aangaan en onderhouden van betekenisvolle relaties met stakeholders (Dumay, 2009, Roobeek, 2005, Waasdorp. 2010). Met de komst van sociale media in het afgelopen decennium wordt in zowel in de wetenschappelijke als semi-wetenschappelijke literatuur in toenemende mate gesproken van een 3.0 organisatie (o.a. Van het Hoff, 2011, Solis, 2012): feitelijk een ‘netwerkorganisatie’, met als toevoeging dat een organisatie een netwerk met stakeholders (leveranciers, concurrenten, klanten e.d.) aangaat en onderhoudt met behulp van sociale media. Soortgelijk spreekt Van Belleghem (2012) van de ´Conversation Company´ en heeft Gray (2012) het over de ‘Connected Company’, waarbij de eerste auteur de nadruk legt op conversaties tussen organisatie en klant, en de tweede auteur op de relaties tussen diezelfde twee. Pilot-onderzoek – opzet In het voorjaar van 2012 is bij zeven organisaties in de sector zorg & welzijn een explorerend onderzoek uitgevoerd. In iedere organisatie zijn diepte-interviews met management en communicatiespecialisten gehouden. Ook hebben focusgroep-bijeenkomsten met medewerkers plaatsgevonden, en is deskresearch uitgevoerd. Resultaten 1. Methodisch/modelmatig resultaat: de indicatoren uit het 7I netwerkmodel beschrijven gezamenlijk de belangrijkste aspecten van een 3.0 organisatie: het model bleek aan te sluiten bij de 160
161
ETMAAL 2013
BOOK OF ABSTRACTS
inrichting van de onderzochte organisaties, intern èn extern. Naar aanleiding van het on-derzoek zijn de indicatoren wel aangescherpt. Daarnaast zijn de structuur en de onderlinge re-laties van de I’s aangepast. Hierbij is een onderscheidende positie toegekend aan de I Identi-teit als centrale indicator. Daaromheen is de I Interne organisatie gemodelleerd; hieronder val-len zaken als: structuur, cultuur en leiderschap. Aan de buitenkant van het model bevinden zich de vier I’s Inzichten, Interactie, Inspirerend netwerk en Innovatie, als onderscheidende aspecten van het extern handelen van een organisatie (met stakeholders). De laatste I In-strumenten heeft vooral een randvoorwaardelijke functie. Zie figuur 1. 2. Inhoudelijk/praktisch resultaat: de meeste organisaties zijn vooral op het gebied van de exter-ne indicatoren (nog) geen 3.0 organisatie. Intern wordt (veelal) aan de belangrijkste rand-voorwaarden voldaan. Zo is er bij de meeste organisaties sprake van een platte/dynamische organisatie, een open cultuur en zelfsturende medewerkers. Ook zijn de benodigde instru-menten in basis beschikbaar (denk aan: PC, tablet, smartphone met internettoegang), en is er sprake van een open houding t.a.v. social media (gebruik onder werktijd). Echter, organisa-ties/medewerkers zetten deze faciliteiten beperkt/niet in om in interactie te treden met stake-holders. Beleid hieromtrent ontbreekt. Dit terwijl men dit offline wel doet (bv. via netwerkbij-eenkomsten / face-to-face contact).
'NU MOET NOG BLIJKEN WAT ZE GAAN DOEN. MAAR HET VOELT POSITIEF!' ONDERZOEK NAAR DE BIJDRAGE VAN BEWONERSBIJEENKOMSTEN AAN HET CONTACT TUSSEN BEWONERS EN DE GEMEENTE ALMERE - CHRISTINE BLEIJENBERG, NICK DESSING AND REINT JAN RENES Organisatiecommunicatie: de bijdrage van organisatiecommunicatie aan het in tune zijn van organisaties met hun omgeving. Relatie burger en overheid: hoe bewonersbijeenkomsten bijdragen aan de relatie tussen burgers en de overheid. Interactief beleid: bewonersbijeenkomsten als instrument voor interactie of van interactief beleid
Session #3 Thursday 7 February 16:45 – 17:45 hrs. Chair: Claartje ter Hoeven Location: T3-34 English/Dutch session
THE COMMUNICATION OF EMOTIONS BY ORGANIZATIONS IN TIMES OF CRISIS: THE EFFECT OF CRISIS RESPONSE STRATEGY AND COMMUNICATED EMOTIONS ON CORPORATE REPUTATION TONI G.L.A. VAN DER MEER AND JOOST W.M. VERHOEVEN An organizational crisis is usually a highly emotional experience for organizations, as well as for stakeholders. Members of organizations in crises are often very engaged in a crises and, as a consequence, are very likely to experience negative emotions during such a crisis. Such emotions may alter the way the members of an organization communicate about a crisis. Alternatively, one 161
162
ETMAAL 2013
BOOK OF ABSTRACTS
may choose to try to suppress ones emotions, and respond in a rational way. Expressed emotions, such as shame and regret, can offer the public valuable cues in interpreting the organization’s role in a crisis. In addition, the organizational expression of emotions is considered to affect the public interpretation, and thus the effectiveness of crisis response messages. Despite the importance of emotions in crisis situations, little is known about the effect of communicated emotion in crisis communication on corporate reputation. 147 Respondents (44% females, Mage = 34,9, SD = 14,5) participated in a online scenario experiment with a 3 (communicated emotion: shame versus regret versus no emotion) x 2 (crisis response strategy: rebuild vs. diminish) factorial design. The fictional crisis concerned the recall of a product (a bike or a car) by an organization. After reading this scenario, participants filled out a questionnaire measuring the anticipated emotions (schadenfreude/anger, sympathy), perceived integrity and account acceptance, and corporate reputation. This scenario experiment revealed that, besides crisis response strategy (diminish vs. rebuild), the communication of emotion (shame and regret) positively affected corporate reputation. The communication of shame and regret proved to be an effective way to protect ones reputation. Mediation analyses showed that this effect of communicated emotion is the consequence of the public’s affective as well as cognitive responses. More precisely, the effect is mediated by positive and negative affect, by account acceptance, and by perceived integrity. Furthermore, a three way interaction effect revealed that the communication of emotions can strengthen the effect of a crisis communication strategy for women, but not for men; women tend to judge an organization which applies a rebuild strategy more positive than an organization which applies a diminish strategy when this is combined with the communication of emotions. Findings confirm that the communication of emotions during a crisis can not only impact reputations directly, but can also impact the interpretation of the crisis response message. The challenge for PR professionals is to find out when to express or suppress what emotions. While the effects of other emotions should still be tested, regret and shame seemed to be appropriate in case of a crisis and an organization should not hesitate to express these authentically felt emotions. Besides developing and implementing an effective crisis response strategy, mastering the communication of ones emotions contributes to the protection of one’s corporate reputation during a crisis. However, the authors do warn against the communication of unauthentic or fake emotions. Such deceptive use of communicated emotions can easily backfire.
NEGATIEVE BERICHTGEVING OVER ÉÉN BEDRIJF BENADEELT DE REPUTATIE VAN DE HELE BEDRIJFSTAK - JOS HORNIKX AND NICOLE VAN DER STEEN Theoretisch kader Een goede reputatie is voor bedrijven van het allerhoogste belang voor het behouden van klanten, het aantrekken van betrouwbare leveranciers en het onderhouden van warme contacten met de media (Walker, 2010). Empirisch onderzoek laat zien dat negatieve berichten over een bedrijf tot een verslechtering van de reputatie leiden (Einwiller, Carroll, & Korn, 2010; Meijer & Kleinnijenhuis, 2006). In modellen van crisiscommunicatie (bv. Coombs, 2007; Coombs & Holladay, 2006) wordt in de regel alleen gekeken naar individuele bedrijven: om welke crisis gaat het, welke factoren spelen 162
163
ETMAAL 2013
BOOK OF ABSTRACTS
een rol en wat kan dat bedrijf het beste doen? Het is mogelijk dat negatief nieuws niet alleen de reputatie van het betreffende bedrijf schaadt, maar ook een bedreiging voor de gehele bedrijfstak vormt (Coombs, 2006). In zo’n geval wordt een groep bedrijven beoordeeld aan de hand van informatie over één bedrijf. Uit onderzoek blijkt dat mensen kenmerken van één object (hier: één bedrijf) in hoge mate generaliseren naar de populatie van objecten (hier: meerdere bedrijven) (bv. Zillmann, 1999; Tversky & Kahneman, 1974). De eerste verwachting is daarom: H1 Negatieve berichtgeving over één bedrijf beïnvloedt ook de reputatie van de gehele bedrijfstak Een generalisatie van één object naar een populatie is eigenlijk alleen verantwoord als dat object in hoge mate representatief is voor die populatie. Dit criterium wordt in de praktijk vaak geschonden: mensen generaliseren ongeacht de representativiteit van het ene object (Gibson & Zillmann, 1994; Hamill, Wilson, & Nisbett, 1980). De tweede verwachting is daarom: H2 Negatieve berichtgeving over één bedrijf beïnvloedt de reputatie van de gehele bedrijfstak, ongeacht de mate van representativiteit van dat bedrijf voor die bedrijfstak Methode Het experiment had een 2 (neutraal – negatief bericht) x 2 (hoge – lage representativiteit) tussenproefpersoonontwerp. Willekeurig kregen 147 proefpersonen (leeftijd: M = 34.54, SD = 16.72; 53.7% vrouw; 44.5% hoger onderwijs) één van de vier teksten voorgelegd. De groepen verschilden niet in gemiddelde leeftijd, verdeling van geslacht en opleiding. Het materiaal bestond uit een nieuwsbericht over de Franse supermarktketen Auchan. De helft van de nieuwsberichten was neutraal van aard: Auchan zou de verpakkingen van versproducten van het huismerk aanpassen. In de andere helft stond vermeld dat Auchan schuldig werd bevonden aan het frauderen met THTdatums. In de achtergrondinformatie over Auchan werd het bedrijf of als zeer vergelijkbaar met andere Franse supermarkten geschetst (bv. weinig tot geen onderscheidende kenmerken) of als een supermarkt met een unieke, onderscheidende positie. De vier condities waren gelijk in opmaak, lettertype en vormgeving. De reputatie van Auchan en van de bedrijfstak werd gemeten met de constructen betrouwbaarheid, deskundigheid en aantrekkelijkheid (O’Keefe, 1990; α’s > .91). In de ene helft van de condities werd eerst Auchan’s reputatie gemeten, in de andere versie de bedrijfstak. Deze volgorde had geen effect op de resultaten. Proefpersonen beantwoordden daarnaast vragen over de begrijpelijkheid en de aard van het nieuwsbericht, de mate van representativiteit, de betrokkenheid van de proefpersoon bij de supermarktketen en de natuurlijkheid van het nieuwsbericht. Resultaten De manipulaties van bericht (neutraal – negatief) en representativiteit (laag – hoog) bleken geslaagd. Een multivariate variantieanalyse liet een hoofdeffect zien van aard van de berichtgeving op de reputatie van Auchan (F (3, 139) = 120.05, p < .001, n2 = .72): Auchan werd minder gewaardeerd na het lezen van het negatieve nieuws. Voor H1 liet een tweede multivariate variantieanalyse een soortgelijk hoofdeffect zien op de reputatie van de bedrijfstak (F (3, 140) = 17.29, p < .001, n2 = .27): de bedrijfstak werd minder gewaardeerd na het lezen van het negatieve nieuws over Auchan. Omdat er geen interactie optrad tussen aard en representativiteit, vond H2 ondersteuning (F (1, 143) = 2.17, p = .14). Het effect van aard van de berichtgeving op de reputatie van de bedrijfstak bleek dus niet af 163
164
ETMAAL 2013
BOOK OF ABSTRACTS
te hangen van de mate waarin Auchan representatief was voor de bedrijfstak. Conclusie en discussie Dit experiment levert een belangrijke bijdrage aan onderzoek naar het effect van negatieve berichtgeving op reputatie door te laten zien dat negatieve berichtgeving ook de reputatie van de bedrijfstak kan schaden. Voor bedrijven en voor onderzoek naar crisiscommunicatie betekent dit dat slecht nieuws over concurrenten ook moet worden meegenomen als factor bij de eigen reputatie.
HOE EEN ORGANISATIECRISIS WORDT GEFRAMED DOOR ORGANISATIE EN PERS: EEN CASESTUDIE OVER DE BP-OLIERAMP IN DE GOLF VAN MEXICO - JULIA KRAUWER AND PYTRIK SCHAFRAAD Dit onderzoek geeft nieuw inzicht in het proces van frame-editing in de context van een corporate crisis, door een vergelijking te maken tussen frames in pers- en krantenberichten. De veelbesproken BP-olieramp in de Golf van Mexico in 2010 diende als casus. In onderzoek naar crisis en risicocommunicatie staat vaak de vraag naar de invloed van de communicatie van organisaties op hun reputatie centraal, wat meestal onderzocht wordt door middel van experimenten. Veruit de meeste informatie over dit soort organisationele crises ontvangen stakeholders nog immer via de nieuwsmedia. Daarom is de inhoudelijke onderzoeksvraag in dit project gericht op de verschillen en overeenkomsten tussen de persberichten van BP en de nieuwsberichtgeving van kranten wat betreft de framing van deze olieramp. Het materiaal voor dit onderzoek bestaat uit alle op de website van BP gepubliceerde persberichten over deze corporate crisis en de nieuwsberichten in drie Britse zogenaamde kwaliteitskranten (Brits vanwege de Britse identiteit van BP). Het onderzoek is uitgevoerd met behulp van een kwantitatieve computer-geassisteerde inhoudsanalyse (co-word analyse). Deze methode is gebaseerd op het voorkomen van (niet vooraf bepaalde) woordcombinaties (die mogelijk een frame representeren). Deze analyse gaat via een aantal stappen waarbij de eerste stap het samenstellen van een basiswoordenlijst is. Deze is de basis voor de verdere stappen in de analyse. In dit type co-word analyse wordt de basiswoordenlijst in principe op basis van woordfrequentie samengesteld. Het is gebruikelijk om daarbij alle z.g. ‘stopwoorden’ vooraf te verwijderen. Echter, dat laat ruimte om verder te selecteren en bovendien inhoudelijke of theoretische selectiecriteria toe te voegen. Dat laatste is in dit onderzoek is de selectie uitdrukkelijk gedaan vanuit het perspectief van theoretische en case-inhoudelijke relevantie, om zo de construct-validiteit van het meetinstrument te verhogen ten opzichte van de zuiver op frequentie gebaseerde selectie. Tot op heden is er in de literatuur over deze methode, zover wij weten, weinig aandacht besteed, aan de samenstelling van deze woordenlijst (anders dan op basis van frequentie). Door expliciet selectiecriteria te formuleren, reduceren we de ambiguïteit van dit belangrijke proces in deze analysemethode. Uit een corpus van meer dan 400 teksten werden zes issue-specifieke frames gedestilleerd. Het update-frame, financiële gevolgen-frame, behoedzaam handelen-frame, internationale crisis-frame, politiek beraad-frame en mens & milieu-frame bleek een centrale rol te spelen in het discours rond de olieramp. Het, met behulp van de netwerkanalyse software van Pajek gevisualiseerde, discours in de persberichten van BP concentreerde zich consequent op het update-frame, terwijl het 164
165
ETMAAL 2013
BOOK OF ABSTRACTS
krantendiscours veel breder en dus meer divers was. Daarin bleek vooral het internationale crisisframe dominant te zijn. De kranten hanteerden dus een duidelijk andere invalshoek dan BP, maar doorweefden hun berichten her en der met informatie uit de persberichten, wat ook blijkt uit de aanwezigheid van het update-frame in de krantenberichten. Tussen de drie kranten vonden we weinig verschillen. Gezien het onderzoek zes statistisch helder onderscheiden en inhoudelijk goed te duiden frames opleverde, mogen we stellen dat de door ons gekozen inhoudelijke selectiemethode voor de basiswoordenlijst geslaagd is en bovendien te prefereren boven een niet inhoudelijk gemotiveerde selectie, zoals tot dusver gebruikelijk in framingonderzoek gebaseerd op co-word analyses. De verschillen en overeenkomsten tussen de twee discoursen lijken er op te wijzen dat organisaties wel in staat zijn om in een dergelijke crisissituatie door consequent het zelfde framepakket te hanteren hun boodschap door te laten dringen tot de nieuwsberichtgeving, maar dat haar rol niet meer kan zijn dan die van ‘co-editor’. Immers ook andere stakeholders proberen invloed uit te oefenen op de berichtgeving en de journalist heeft als regisseur met een eigen belang (bij voorbeeld gebalanceerde berichtgeving) het laatste woord, waardoor de krantenberichtgeving, zoals verwacht een veel diverser framepaket tentoon spreid, dan de persberichten van BP. Daarbij valt op de er tussen de kranten onderling weinig verschil gevonden werd; elk van de kranten geeft dus een divers beeld van de gebeurtenis, maar deze diverse beelden lijken sterk op elkaar wat duidt op een geringe mate van pluriformiteit in het ‘kwaliteitskrantensegment’.
DE COMMUNICATIEPROFESSIONAL ALS BOUNDARYSPANNER; EEN STUDIE NAAR DE OMGEVINGSCOMMUNICATIE RONDOM DE AANLEG VAN DE NOORD/ZUIDLIJN IN AMSTERDAM NOELLE AARTS AND MIEKE MUIJRES De Noord/Zuidlijn zal Amsterdam-Noord met Amsterdam-Zuid verbinden met een reistijd van 16 minuten. De lijn is 9,7 kilometer lang en krijgt acht stations. Vanwege negatieve ervaringen met de aanleg van de Oostlijn lag het idee van een tweede metrolijn in de stad van begin af aan gevoelig. Na veel overleg stemde de Tweede Kamer in 1999 in met de aanleg van de Noord/Zuidlijn. Uiteindelijk nam de gemeenteraad van Amsterdam in 2002 het definitieve besluit en zijn de werkzaamheden in april 2003 gestart. Door de jaren heen is veel tot stand gebracht, maar is het project ook met forse tegenvallers geconfronteerd. Medio 2009 bereikte het imago van de Noord/Zuidlijn een absoluut dieptepunt. De kostenbegrotingen werden meerder malen naar boven bijgesteld en de ingebruikname van de metroverbinding werd noodgedwongen uitgesteld van 2011 naar 2017. Bovendien verzakten bij de werkzaamheden aan het diepe station Vijzelgracht verschillende oude panden, waaronder de monumentale Wevershuisjes aan de Vijzelgracht. Deze gebeurtenis betekende een belangrijk omslagpunt in de beeldvorming rondom de Noord/Zuidlijn. Het vertrouwen in het project en de projectorganisatie wankelde doordat valse beloften waren gedaan over de kosten en oplevering, en door de gebrekkige communicatie over risico’s en overlast voor de omgeving. In 2009 riep het Amsterdamse College van Burgemeester en Wethouders een onafhankelijke commissie, onder leiding van oud-minister Veerman, bij elkaar om te onderzoeken of, en zo ja, hoe het verder moest met de Noord/Zuidlijn. Het antwoord op deze vraag van de commissie was duidelijk: de Noord/Zuidlijn moest zo snel mogelijk worden afgemaakt. De commissie stelde echter wel een aantal 165
166
ETMAAL 2013
BOOK OF ABSTRACTS
belangrijke voorwaarden op die nageleefd moesten worden. Belangrijk onderdeel van die voorwaarden was een betere communicatie met de omgeving. De commissie Veerman had het over ‘een gebrek aan empathie’ en noemde het Noord/Zuidlijn project een ‘autistische organisatie’ met te weinig oog voor de omgeving. Op signalen uit de omgeving werd, als ze al werden opgepikt, niet adequaat gereageerd. In plaats daarvan gedroeg de organisatie zich zelfreferentieel: men ging uit van een volstrekt eigen idee over wat belangrijk was en wat niet. Het begrip zelfreferentialiteit heeft enige overlap met het begrip selectieve waarneming: mensen filteren uit hun omgeving de informatie die op een of andere manier nuttig lijkt en die niet als bedreigend wordt ervaren. Aldus ziet men bepaalde dingen en blijven anderen onzichtbaar. Het probleem van zelfreferentialiteit is dat daardoor op een zeker moment de aansluiting met de omgeving wordt gemist en het bestaansrecht van een organisatie in het geding komt. De nieuwe communicatieopgave was gericht op het doorbreken van de neiging tot zelfreferentialiteit. De organisatie moest zichzelf leren zien door de ogen van de omgeving. Alleen op die manier worden signalen opgepikt die vervolgens tot actie leiden waarmee afstemming met de omgeving tot stand komt. In dit paper doen we verslag van een studie naar de wijze waarop de communicatie rondom de NoordZuidlijn vanaf dat moment is aangepakt en hoe we het resultaat daarvan theoretisch kunnen duiden. Als basis voor interviews met belanghebbenden bij de bouw van de NoordZuidlijn is een theoretisch kader ontwikkeld waarin de concepten ambitie, interactie en omstandigheden centraal staan. De resultaten van onze studie laten zien dat de communicatierondom de NoordZuidlijn gekenmerkt wordt door een grote nadruk op het voortdurend organiseren van verbindingen tussen belanghebbenden. Op basis van gesprekken, een open, luisterende houding en het organiseren van oplossingen van problemen van betrokkenen zijn relaties hersteld en is vertrouwen gecreëerd. De resultaten laten ook zien dat een verbrede visie op de rol van communicatie bij complexe en relatief onvoorspelbare processen zoals de aanleg van de NoordZuidlijn vruchten afwerpt. Daarbij ligt de nadruk op voortdurende interactie en een niet aflatend oog voor de omstandigheden. De ambities gaan verder dan het zenden van boodschappen: de communicatie is ook, en misschien wel vooral, gericht op het creëren van oplossingen voor problemen die zich voordoen. Daartoe is een tweezijdige opvatting van communicatie noodzakelijk. Geconcludeerd wordt dat ambities uiteindelijk worden gerealiseerd in gesprekken met betrokkenen, waarbij onverwachte omstandigheden niet alleen als obstakel naar voren komen, maar ook als opportunities. Om deze opportunities te ontdekken is een flexibele houding van communicatieprofessionals vereist, waarbij een hoge mate van onvoorspelbaarheid wordt geaccepteerd. Een strakke planning van communicatie op basis van vooraf gestelde doelen wordt hiermee onmogelijk. In plaats daarvan is succesvolle communicatie rondom complexe opgaven vooral het resultaat van een open, luisterende houding die gepaard gaat met transparantie over succes én falen.
166
167
ETMAAL 2013
BOOK OF ABSTRACTS
Session #4 Friday 8 February 09:00 – 10:00 hrs. Chair: Vidhi Chaudhri Location: Auckland (M3-04) English session
THE CHANGING ORGANIZATION OF HUMANITARIAN INFORMATION PROVISION: CHALLENGES OF TECHNOLOGY-MEDIATED CITIZEN AND VOLUNTEER PARTICIPATION - ANNEMIJN VAN GORP Natural and manmade disasters in developing countries have become more frequent and complex throughout the 1990s. As a result, coordination of relief operations by governmental, private sector and non-governmental organizations has become more complex as well, where particularly the influx of many international aid agencies poses significant challenges. Assessing emergency shelter, health, and sanitation needs of the affected community requires vast amounts of information to be gathered as well as to be shared within the relief community in order to provide a coordinated, effective response. Whereas the increased availability of information and communication technologies (ICTs) has technically eased information distribution, changed patterns of information acquisition have posed new challenges over the last few years. This was particularly evident during the Haitian relief effort, where citizens used social media such as Twitter on a wide scale to disseminate requests for help. Of course, members from the affected community know their surroundings best and are thus arguably in the best position to identify and communicate needs. Yet, the vast amounts of data floating around on social media could be false or misinterpreted. Hence, data need to be filtered, verified and organized. So-called Volunteer and Technical Communities (V&TCs) have taken efforts to aggregate this information and for example plot maps to support aid organizations in their relief efforts. Not only have volunteers provided support in aggregating information. They have also developed technologies that enable easier sharing and filtering of citizen information, such as SMS-based technologies that make information sharing easier when internet connectivity is lacking, or for instance software that helps reunite separated families through registration of missing and found people. The rise of citizen and VT&C participation in humanitarian relief brings new coordination challenges to the traditional mix of state, private sector, and nongovernmental organizations. Questions arise such as how do these new participants in the humanitarian community organize information exchange? In what ways do citizens and volunteers observe participate and in which contexts does information exchange arise? How can traditional aid organizations take advantage of the information that these new actors add to the scene? Which organizations do cooperate with V&TCs? And how sustainable and reliable are these V&TCs that often have limited financial means? Answers to these questions are needed for aid organizations to more effectively coordinate their efforts with these communities and to be able to fully take advantage of the services they provide.
167
168
ETMAAL 2013
BOOK OF ABSTRACTS
This study aims to provide some first insights into these questions. The study will start with an overview of the often used ICTs during disaster response, with a particular focus on social media and crowd sourcing technologies. Next, the study will identify and analyze V&TC-based initiatives such as the Standby Task Force and CrisisCommons, and their role in recent relief efforts. In doing so the study will provide insight into the benefits and challenges of V&TCs in humanitarian information provision and the effects of these new stakeholders on the organization of the humanitarian community at large.
NEW WAYS OF WORKING, VITALITY, AND EMPLOYEE PERFORMANCE: THE ROLE OF SELF DETERMINATION - CLAARTJE TER HOEVEN, FREDERIQUE SMIT AND MARIEKE FRANSEN Communication technologies continuously expand the possibility for redesigning the work environment. Organizations often explore these possibilities, due to its potential to create an inspiring and innovative workplace (Clercq, Bulent, & Auh, 2008), increase work flexibility (Gajendran & Harrison, 2007), reduce organizational costs (Rennecker & Godwin, 2005), and create competitive advantage (Rau & Hyland, 2002). Therefore, many employees (are asked to) work with communication technologies and under flexible circumstances (Galinsky, Bond, & Sakai, 2008; OECD, 2011). Work designs characterized by temporal (schedule flexibility) and spatial (telecommuting) flexibility and facilitated by communication technology are referred to as ‘new ways of working’ (NWW; Ten Brummelhuis, Bakker, Hetland, & Keulemans, 2012). A quickly expanding body of literature demonstrates the positive outcomes of employing NWW (Gajendran & Harrison, 2007; De Menezes & Kelliher, 2011). However, it is not clear which underlying process causes these positive work outcomes and if these outcomes are the same for all employees. We argue that flexible work designs generate productive work behavior, through experienced vitality while at work. NWW facilitates the possibility to combine work and private life (Greenhaus & Powell, 2006) and may save troubles and costs by diminishing commuting time. This could mean that employees have energy to spare. The remaining energy can then be used to work more efficient and effective, in other words, to be more productive. Therefore, the first aim of this study is to examine the mediating role of vitality in the NWW-performance link. Furthermore, employees may differ in their motivation and ability to alter their job design (Grant & Parker, 2009). The amount of self-determination is of special interest in the NWW context, since employees who are highly self-determined are known to operate autonomous and proactive. These characteristics might help to fully benefits from the flexibility offered by NWW. The second aim of this study, is, therefore, to test the assumption that selfdetermination strengthens the link between NWW and the work outcomes vitality and productivity, in such a way that the link between NWW and the work outcomes is more pronounced when selfdetermination is high compared with when self-determination is low. Data were collected at a Dutch international financial firm, using a web-based questionnaire. A total of 830 employees received the questionnaire and 324 were returned (response rate 39%). The hypotheses were tested by conducting structural equation modeling (SEM) using AMOS (Arbuckle, 2006). SEM is a preferable data analysis strategy for mediational models involving latent constructs (Baron & Kenny, 1986). We used bootstrapping to test whether the significant pathways running between the NWW dimensions and performance via vitality do in fact represent mediated relationships. To test the interaction effects of self-determination we conducted MSEM analyses, using AMOS (Arbuckle, 2006), as MSEM allows for assessing and correcting for measurement error 168
169
ETMAAL 2013
BOOK OF ABSTRACTS
whereas it provides measures of the fit of the model under study at the same time. The measurement model included the scale variables of temporal flexibility, spatial flexibility, connectivity, self-determination, vitality, and performance. This model showed an acceptable fit to the data, 2(281) = 535,17, RMSEA = .053, GFI = .89, CFI = .92, TLI = .91. The structural model includes five latent constructs: temporal flexibility, spatial flexibility, connectivity, vitality, and performance. This model showed a good fit to the data (2(198) = 386,89, RMSEA = .054, GFI = .91, CFI = .94, TLI = .92). The bootstrapping results indeed showed a positive association between the subdimensions of NWW and employee performance mediated through vitality. Furthermore, the MSEM analyses revealed a moderating effect of self-determination, in the expected direction. The first objective of this study was to examine the mediating role of vitality in the NWW-performance link. We indeed found support for this proposition. It appears that a flexible work design facilitated by communication technology enhances employee performance through experienced vitality at work. The second aim of this study was to examine the moderating role of self-determination in the relationship of NWW with vitality and employee performance. The results of this study supported this assumption. This underlines the importance of autonomous and proactive employees to fully benefit from the possibilities offered by NWW. In other words, this work design will be especially fruitful for certain employees whereas others might need more support by rearranging their (work) life according to these new circumstances.
ONE SIZE FITS ALL? ADVANTAGES AND CHALLENGES OF NEW WAYS OF WORKING: THE ROLE OF PARTICIPATIVE LEADERSHIP - PERNILL VAN DER RIJT, CLAARTJE TER HOEVEN, JOEP HOFHUIS AND MARTIJN VLUG-MAHABALI In recent decades, the organization of work has changed substantially. It used to be organized in a central location and took place during fixed hours; now, the work process is mostly flexible. This flexibility has been enabled by communication technologies. These flexible work designs are referred to as New Ways of Working (NWW). In contrast with the quickly expanding body of literature demonstrating the advantages of employing NWW (Gajendran & Harrison, 2007), the challenges employees and supervisors may face when using this job design are examined less extensively. Recently, however, some challenges of NWW surfaced in the literature. The conflicting findings about the effects of NWW raise a number of questions. First, which of the advantages and which of the challenges associated with NWW affect employee performance and work satisfaction? Second, under what circumstances do these advantages and challenges occur? This study aims to answer these questions, extending previous research in at least two ways. First, we investigated two underlying mechanisms that link NWW and performance. Previous studies on flexible work arrangements and employee performance have resulted in mixed findings ranging from positive association to no significant relationship. The existence of both advantages (e.g., job control) and challenges (e.g., exhaustion) associated with NWW could account for the none findings in some of these studies. In other words, in multiple mediator models, it is not always possible to find a significant direct relationship between x and y because two mediator variables (m1 and m2) can have opposite effects that outweigh the net effect of x on y . Therefore, a multiple mediator model is proposed here to test the effects of the advantage job control and the challenge work exhaustion on the relationship between NWW and employee performance. 169
170
ETMAAL 2013
BOOK OF ABSTRACTS
Second, an explanation of the mixed findings might be the conditions under which NWW has an effect. Employees may be enthusiastic about flexible work policies, but if the leaders do not support such policies, they are less likely to meet with success. Managerial support for this work design may also vary because it may reduce managers’ control and limit transparency about how and when work is completed. Therefore, managerial support (e.g., participative leadership) is taken into account as moderator. Multiple source data was collected in 2012 with 159 pairs: 159 employees filled in a questionnaire about, for example, their use of the job design NWW, perceived job control, work exhaustion, and job satisfaction. Their manager filled in a questionnaire about the job performance of the employee. The proposed model was tested with structural equation modeling (SEM) using AMOS. Bootstrapping was used to test whether the significant pathways running between NWW and job satisfaction and performance do in fact represent mediated relationships. To test the interaction effects of participative leadership we conducted MSEM analyses, using AMOS (Arbuckle, 2006), as MSEM allows for assessing and correcting for measurement error whereas it provides measures of the fit of the model under study at the same time. The bootstrapping results indeed showed a positive relationship between NWW and the work outcomes job satisfaction and performance, with a mediating effect for job control. We did not find a mediation effect for work exhaustion. Furthermore, the MSEM analyses revealed a moderating effect of participative leadership, in such a way that the association between NWW and job control was stronger for employees who had a leader who scored high on participative leadership as compared to low. The first objective of this study was to examine a multiple mediator model (e.g., job control and work exhaustion) in the NWW-performance/job satisfaction link. We only found support for the mediation of job control in this relationship. It appears that a flexible work design facilitated by communication technology enhances employee performance through experienced job control. The second aim of this study was to examine the moderating role of participative leadership in the relationship of NWW with perceived job control. The results showed that participative leadership is an important leadership style, in the context of NWW.
Session #5 Friday 8 February 10:15 – 11:15 hrs. Chair: Piet Verhoeven Location: Auckland (M3-04) English/Dutch session
INFORMATION SHARING AS STRATEGIC BEHAVIOUR: THE ROLE OF SOCIAL MOTIVATION, TIME PRESSURE AND INFORMATION DISPLAY - NICOLETA BĂLĂU AND SONJA UTZ In our knowledge economy, organisations are more and more dependent on employees’ information and knowledge sharing behaviour. Knowledge is an intangible asset (Bontis, Dragonetti, Jacobsen, & Roos, 1999) resulting in a competitive advantage. More and more work is conducted in teams and 170
171
ETMAAL 2013
BOOK OF ABSTRACTS
information sharing is a precondition of high decision quality (Wittenbaum & Park, 2001), a fundamental knowledge-centered activity. Consequently, organizations invest in technology to stimulate organizational knowledge and information sharing. Whereas literature from the domain of information or management systems mainly focuses on the characteristics of knowledge management systems or social media tools, literature from organization science has focused on organizational culture, the role of leadership, or communication styles (de Vries, van den Hooff, & de Ridder, 2006). However, although it has been mentioned that knowledge sharing is a social dilemma and represents a mixed-motive situation (Cabrera & Cabrera, 2002), empirical research has largely neglected the strategic aspects of information sharing. Steinel, Utz, and Koning (2010) showed in a series of experiments that social motivation plays an important role. Pro-self individuals are often not motivated to share all their knowledge; they share some pieces of information to make a cooperative impression but keep the important information for themselves. The present study enhances the understanding of strategic information sharing in three ways. Firstly it aims to replicate the effect of social motivation on strategic information sharing reported by Steinel et al. (2010), in a more complex setting involving actual and more (28 instead of 12 pieces of) information. It’s being hypothesized that pro-socials share more important information (H1a) as well as more private information (H1b) than pro-selfs. Secondly, this paper builds on this line of research and extends it by examining also the role of need for cognitive closure (NFCC) (Webster & Kruglanski, 1994). NFCC is a person’s motivation with respect to information processing and judgement; in this study it will be manipulated by varying time pressure. Individuals with a high NFCC search for less information and might therefore also share less information. Specifically, it’s being hypothesized that individuals under high NFCC share less important (H2a) and less private (H2b) information than individuals under low NFCC. This second aim becomes timely as many decisions in organizations have to be made under time pressure. Thirdly, we aim to study whether the display of information affects information sharing behaviour by comparing a situation in which information has to be pulled from different folders with a situation in which all information is pushed directly to the individuals. This is a relevant purpose since RSS or twitter feeds present people with large amounts of information on a specific topic. These technologies provide a more efficient access to information, but might also create a situation of information overload. The way how information has been accessed might influence how readily people are willing to share it. An open research question is formulated: (RQ) How is information sharing behaviour affected by the push/pull information display? Method Students (N = 125: 53 males, 72 females; age: M = 22.3, SD = 5.22) from VU University Amsterdam participated in the study. A 2 (social motivation: pro-social vs. pro-self) x 2 (need for cognitive closure: high vs. low) x 2 (information display: push vs. pull) experimental design was used; social motive, NFCC and information display were varied between participants; sharedness and importance of information were varied within participants. Results In line with previous findings (Steinel et al., 2010), pro-socials share significantly more important information than pro-selfs (H1a supported). A marginal effect showed that pro-socials tend to share more private information than pro-selfs (H1b partially supported). Although no support was found for H2a, a marginal effect showed that people under high NFCC tend to share less private 171
172
ETMAAL 2013
BOOK OF ABSTRACTS
information than people under low NFCC (H2b partially supported). This has important implications for organizations because it is especially important to increase the sharing of private information. A significant main effect of NFCC showed also that, overall, people under high NFCC shared less information than people under low NFCC. This is to our knowledge the first study demonstrating that NFCC does not only influence information processing but also information sharing. For our RQ, the main effect of information display on information sharing was not significant. However, we detected a complex interplay between information display, information sharedness and social motive: prosocials shared significantly more private information than pro-selfs in the push information display condition. In addition, also in the same push condition, pro-socials shared significantly more private than public information.
DO MEDIA MATTER IN ORGANIZATIONAL IDENTIFICATION? THE ROLES OF MEDIA USAGE AND PERCEPTION OF COMMUNICATION IN EXPLAINING EMPLOYEE IDENTIFICATION - D.J. SCHUTTEN, J. BARTELS AND J.W. OUWERKERK In the past decades organizational identification has been a strong concept in explaining positive outcomes (Ashforth et al., 2008). Employees who identify strongly with their organization have more positive attitudes towards the organization (Mael & Ashforth, 1992), have less intention to leave (Van Dick et al., 2004), are more satisfied with their job (Van Knippenberg & Van Schie, 2000), show more organizational citizenship behaviour (Bergami & Bagozzi, 2000) and cooperate more intensively with other organizational members (Dutton et al., 1994). Although organizational identification has been studied in a variety of contexts, specifically organizational communication has proven to play an important role in explaining employees’ identification with or commitment to the organization (Bartels et al., 2007; Smidts et al., 2001; Postmes et al., 2001). Smidts et al. (2001) found that employees’ perceptions of communication mediated the effect of organizational information received on their organizational identification. Bartels et al. (2007) first described the influence of communication climate on employees’ identification with different organizational levels. In an organizational commitment context, Postmes et al. (2001) found that horizontal communication (informal, with proximate colleagues, of socio-emotional content) is less strongly related to levels of commitment at both organizational and unit level, than is vertical communication (strategic information and communication with management). Although several studies on the influence of communication on employees’ identification have emerged over the last years, the interplay between perceptions of horizontal and vertical communication and (new) media, and its influences on different forms of identification has hardly been a topic of investigation so far. Since the usage of new media in organizational communication has risen enormously over the last few years (Scott & Timmerman, 2005), the current study aims to investigate the roles that both (new) media and perceptions of vertical and horizontal communication play in explaining employees identification with their organization. We conducted a pilot study among employees in 11 different branches (n = 94). Besides demographics, respondents filled out a self-administered questionnaire on three topics, 1) media usage and evaluation, 2) perceptions of horizontal and vertical communication and 3) departmental and organizational identification. For media usage, respondents were asked to rate several media for communication with their supervisor (vertical use of the medium) and with their colleagues (horizontal usage of the medium). We used social media, face-to-face communication, email and 172
173
ETMAAL 2013
BOOK OF ABSTRACTS
telephone as media. Perceptions of vertical and horizontal communication were both measured with a five-item scale based on Bartels et al. (2010) and Dennis (1974). CFA showed that a twodimensional model (vertical and horizontal communication as separate constructs) provided a much better fit (х2/d.f. = 2.67; GFI = .94; TLI = .92) than a one-dimensional model (х2/d.f. = 13.40; GFI = .56; TLI = .44). Cronbach’s alphas were .96 for vertical communication and .93 for horizontal communication. Departmental and organizational identification were both measured using a sevenitem scale by Leach et al. (2008). Cronbach’s alphas were .96 for departmental identification and .95 for organizational identification. The first results show that face to face communication with one’s supervisor (ß =.19; p <.05), the perception of top-down communication (ß =.47; p <.01) and the perception of horizontal communication (ß =.18; p <.05) are significantly related to employees’ organizational identification. Employees’ identification with the department is only significantly influenced by the perception of top-down communication (ß =.46; p <.01). The usage and evaluation of other media (social media, email and telephone) do not seem to have a direct effect on organizational or departmental identification. In line with Wiesenfeld et al. (1999) there seem to be hardly any direct effects of media usage and media evaluation on employees’ identification. Of all the media taken into account, only usage and evaluation of face to face communication with one’s supervisor has an effect on employees’ organizational identification. Moreover, the results of the current study seem to confirm earlier results by Bartels et al. (2010), who found that vertical communication significantly affected organizational identification. The main finding of the current study is that employees’ perception of top-down communication is most strongly related to employees’ identification with both the department and with the organization. This effect remains, even when including horizontal and vertical media usage and evaluation into the equation. In sum, although the usage of new media in organizations is on the rise, they seem to be less important for employees’ identification than perceptions of top-down communication.
PROFESSIONALIZATION OF ORGANIZATIONAL COMMUNICATION BY MEANS OF DECISION SUPPORT SYSTEMS - MAARTEN C.A. VAN DER SANDEN, KOEN H. VAN DAM AND JEROEN STRAGIER [Introduction] Corporate communication professionals make strategic decisions in highly complex, dynamic and uncertain contexts. These decisions are based on the professional’s knowledge (mostly implicit), experience, intuition and creativity. Despite the fact that these decisions might often be effective, they contain a high level of uncertainty. In the context of professionalization of organizational communication this uncertainty needs to be addressed and made explicit, leading to insights in how to improve one’s own decisions as well as to improve expectation management to other organizational domains. Existing tools, protocols and planning schemes fall short because they lack clear insights in uncertainty levels and moreover, the dynamics of the processes. [Proof of concept] 173
174
ETMAAL 2013
BOOK OF ABSTRACTS
In this paper we show a proof of concept of a simulation based decision support system for corporate communication professionals dealing with the uncertain and dynamic introduction of smart energy grids and smart home energy management. Based on various strategic communication scenarios we make clear how this kind of simulation could support an organizational communication decision dealing with consumer intentions to purchase a smart home energy management system. More specifically, a consumer behaviour model based on the Model of Technology Acceptance in Households (MATH) was quantitatively validated through survey research and used as input data for an agent based model used for simulations. The simulations contain agents that represent households following the technology acceptance model to make decisions on their intent to adopt the new technology. This intent to adopt is influenced by variable communication inputs such as usability of the home energy management system as well as the consumer’s opinion within their own social network. [Results] Results of these simulations illustrate what the effects of, for example, an increase in cost or/and changes in the product usability are on consumer intentions. Some agents show a much slower change of intention than other agents when only the costs increase, whilst they show a quicker change of intentions when in the combination of change in usability and costs is simulated. The same holds for a communication scenario in which changes in the social network of the consumer has a large effect on the consumer’s intention. Some agents show a steep increase in behavioural intent, while others react much slower. When combination effects are simulated, such as change in social network and usability, the curves on intention increase or drop dramatically or curves were offset by social network changes. On the level of simulated neighbourhoods, one sees that the overall neighbourhood intentions sometimes change dramatically. At one moment consumers in a neighbourhood show high intentions, while at another moment the same neighbourhood is less high intentional due to social network changes. [Impact on organizational communication] The simulated results described above show the final level of consumer behaviour intention and the dynamic aspects of that change. Concerning organizational communication, based on these results it becomes more insightful whether such an effect can be changed by communication, in what timeframe and where to best focus such an effort? Where should we start our first in real pilot test? A communication professional can now assess at what neighbourhood, group of consumers, he should focus first. Or it might be the case that the product itself should be modified since the effect of usability change leads to impressive changes in behavioural intent. Or he might decide to ask a strategic alliance to kick off first with their communication campaign on, for example, a sustainable world. A decision support system containing these kind of simulation possibilities makes communication scenarios insightful, and that means that communication professionals can manage expectations towards other organizational management domains more clearly. These systems do not prescribe what to do, but elicit the communication professional in reflecting on what kind of scenarios are possibly useful. [Professional accountability] The use of decision support systems based on simulation may increase the accountability of communication professionals since they are based on simulations incorporating real data and 174
175
ETMAAL 2013
BOOK OF ABSTRACTS
founded in evidence from social sciences and communication literature. As such the communication professional gets support in comparing options for communication processes by estimating their impact under well-defined conditions and scenarios, taking advantage of previous studies so the professional decision no longer solely relies on implicit knowledge, intuition, experience and creativity. Agent-based simulation make uncertainty explicit and generates insights in dynamic and uncertain communication processes, that lead to insightful decisions on communication policy and strategy and results in improved organizational communication.
REMINDERS OF DEATH INFLUENCE ATTITUDE, INTENTION AND BEHAVIOR IN GROUP FORMATION. A TERROR MANAGEMENT PERSPECTIVE ON JOINING A NEW GROUP - JOLIEN ARENDSEN AND JESSICA SCHUTTEN The group is a fundamental and universal feature of human social and professional life, and so group behavior is a widely studied topic. One of the theoretical paradigm that is used to explain group behavior is the Terror Management Theory (TMT; Greenberg, Pyszczynski, & Solomon, 1986). Terror Management Theory suggests that both culture-and group-related behavior are ultimately rooted in the self-preservation motive. To protect oneself against the deeply rooted fear of mortality, one endorses a cultural worldview that gives meaning, order and permanence to oneself (Greenberg, Solomon, & Pyszczynski, 1997). Hundreds of studies support the Terror Management Theory and show that mortality reminders can lead to in-group bias, a variety of nationalistic preferences and discrimination of out-group members (for meta-analysis see: Burke, Martens, & Faucher, 2010). TMT based studies so far have focused on factors concerning groups that one is already a member of. However, in real life one sometimes has to jump in the deep end and join others one doesn’t know, for instance when organizations or departments merge or when combined expertise of different people is needed. And so we wonder what the effects on mortality salience are on joining new groups. To answer this question we conducted an experiment (mortality vs. control) in which participants (N = 98) were all students (24 male, 74 female), with the average age of 25.2 years (SD=3.95). Mortality Salience was evoked by asking the participants to answer two questions about their own death (Greenberg et al., 1990), while participants in the control group were asked to answer two questions about their activities during the day (based on Wakimoto, 2006). After a manipulation check to control for affect and death related thoughts, participants received information about a fictitious project in which they could join a team to compete against other teams from different disciplines. We measured attitude towards the new team (5 item scale) intention to join the team(yes/no question) and actual joining behavior (yes/no question). Results showed no effect of experimental condition (i.e., mortality salience vs. control) on affect: these findings rule out affect as a possible explanation for the effects of mortality salience. A manipulation check showed that participants in the mortality salience condition completed more death related words than participants in the control group: these findings suggest the manipulation in fact evoked death related thoughts. A one way ANOVA showed a significant effect of experimental condition on attitude towards a new team, F(1, 98)=4.63, p<.05, ηp2=.05. Participants in the mortality salience condition had a more negative attitude towards joining a new team (M=3.78 SD=.59) than participants in the control group 175
176
ETMAAL 2013
BOOK OF ABSTRACTS
(M=4.0, SD=.83). A Chi-square test showed a significant effect of experimental condition on the intention to join a new team χ2(2)=5.54, p<.05, =.24. In the mortality salience condition 8.2% of the participants had the intention to join the new team, while in the control condition 26% of the participants wanted to join the new team. In addition, A Chi-square test showed a significant effect of experimental condition on actual joining behavior, χ2(2)=3.87, p<.05, =.20. In the mortality salience condition 4.1% of the participants joined the new team, while in the control condition 16% of the participants joined. Results show a consistent negative effect of mortality salience on attitude towards the new group, intention to join and actual joining behavior. These results are surprising because mortality salience increases the need to share ones values and beliefs (Greenberg et al., 1997). An explanation may be the anonymity of the new group: it is unknown if the new group members share the same cultural values and contribute to ones feelings of worth. These results might be related to the findings of Harmon-Jones et al. (1996) who found that after mortality salience the expected intergroup bias did not occur when persons were randomly assigned to groups. The current study extends previous TMT research on group behavior, by showing that when it comes to the unknown, one doesn’t find affirmation from a new group. This can have implications for all sorts of group forming. In a world where there are constant reminders of death and one is constantly forced to rely on unknown others, this can have great consequences.
176
177
ETMAAL 2013
BOOK OF ABSTRACTS
PERSUASIVE COMMUNICATION Session #1 Thursday 7 February 14:05 – 15:05 hrs. Chair: Karolien Poels Location: Rochester (M2-10) English session
DOES SPONSORSHIP DISCLOSURE TIMING INFLUENCE VIEWERS’ PROCESSING OF PROGRAM CONTENT? - SOPHIE BOERMAN, EVA VAN REIJMERSDAL AND PETER NEIJENS Recent EU regulations require broadcasters to disclose sponsored content in television programs, to inform audiences and guarantee fair communication (Audiovisual Media Services Directive, 2010). This study explores whether sponsorship disclosure makes viewers process the sponsored program content more critically. To critically process sponsored content, people need to be aware of the persuasion attempt (Buijzen, Van Reijmersdal, & Owen, 2010). Sponsorship disclosure aims to achieve this. Therefore, an important antecedent to critical processing might be the activation of persuasion knowledge. Persuasion knowledge is the set of theories and beliefs about persuasion and its tactics that people develop throughout their lives (Friestad & Wright, 1994). Furthermore, when viewers process the program content critically, this may influence the persuasive effect of the sponsored content. Critical processing of the sponsored content may lead to resistance to the brand that it is promoting. Hence, the activation of persuasion knowledge and the subsequent critical processing could lead to less favorable brand attitudes. Additionally, when exploring the effects of disclosure on processing, it is important to consider the timing of disclosure. The moment at which a disclosure is shown in a program determines whether it primes the sponsored content and the opportunity a viewer has to (critically) process the content. This research compares the effects of three different timings: disclosure prior to, concurrent with and after the sponsored content. We propose that a disclosure prior to the sponsored content works as a prime and gives viewers time to critically process the content. A disclosure concurrent with the sponsored content directly emphasizes the sponsored content, which also enables the viewer to process it critically. A disclosure after the sponsored content does not prime nor emphasizes the sponsored content. Therefore, we expect that a disclosure prior to or concurrent with the sponsored content has a larger effect than a disclosure after the sponsored content. In sum, we propose a model (see Figure 1) in which sponsorship disclosure activates persuasion knowledge, which leads to critical processing of the sponsored content, and subsequently negatively affects the attitude toward the brand in the sponsored content. Additionally, we expect that a disclosure prior to or concurrent with the sponsored content has a larger effect than a disclosure 177
178
ETMAAL 2013
BOOK OF ABSTRACTS
after the sponsored content. Method In an experiment, 209 students (77% female, Mage = 22.21, SDage = 3.36) were asked to watch an edited episode of MTV Was Here. They were randomly assigned to one of four conditions: a sponsorship disclosure (“This program contains advertising by Alive Shoes”) was displayed at three different times (prior to, concurrent with or after the sponsored content), or the program did not contain any sponsorship disclosure (control group). The program consisted of three items, and the middle item included the sponsored content as it discusses a new brand of sneakers. Results All participants who were exposed to a disclosure but did not recall this (n = 93) were excluded from analyses. The model was tested using Hayes’ PROCESS macro (Hayes, 2012). Bootstrapping showed significant indirect effects via the two mediators for both the prior (Indirect effect -.11, SE = .07, 95% BCCI [-.302, -.016]) and concurrent condition (Indirect effect -.08, SE = .06, 95% BCCI [-.257, -.006]) compared to the control condition. There was no significant effect of a disclosure shown after the sponsored content (compared to no disclosure; Indirect effect -.04, SE = .05, 95% BCCI [-.184, .024]). Discussion This study demonstrates that sponsorship disclosure leads to persuasion knowledge, which increases critical processing and ultimately leads to less favorable brand attitudes, but only when the disclosure is displayed prior to or concurrent with the sponsored content. Theoretically, the findings emphasize the importance of disclosure timing. A disclosure displayed prior to or concurrent with the sponsored content primes the sponsored content and provides sufficient processing time to viewers to recognize the content as advertising and process it critically. In addition, the findings show that persuasion knowledge and critical processing are important underlying mechanisms for the effect of sponsorship disclosure on brand attitude. Regarding the practical implications for legislators and advertisers, this research demonstrates that sponsorship disclosure can make viewers aware of the sponsored content in television programs. Furthermore, this changes the processing of sponsored content and can also ultimately lead to resistance against persuasion.
SPONSORING GOOD CAUSES: DOES THE LEVEL OF BRAND/CAUSE FIT MATTER? - SIMON ZEBREGS, BAS VAN DEN PUTTE AND MARJOLEIN MOORMAN The value of corporate sponsorship of good causes in North America has increased from 120 million dollar in 1990 up to 1.68 billion dollar in 2011. An important strategy within the field of corporate cause sponsorship is cause-related marketing (CRM). When companies engage in CRM they link the purchase of a product to a donation to a cause. For example, Pampers is running a campaign in which they promise to donate a tetanus vaccination to Unicef for every package of diapers they sell. It has been suggested that the perceived level of brand/cause fit moderates the effects of CRM messages on the brand attitude. That is, when the level of fit is high, people would apply their cause attitude, which is likely to be favorable, to the brand in order to make a judgment about it. However, earlier studies on this topic showed mixed results. Theory on bias correction may explain these results, because it suggests that the positive effect is 178
179
ETMAAL 2013
BOOK OF ABSTRACTS
only likely to occur when people use little cognitive effort while making their judgment. When people use a lot of cognitive effort and are aware that their judgment of the company could be affected by their judgment of the cause, they may correct for this influence. The amount of cognitive effort people use depends on their involvement with the subject. In addition, literature on the effects of information processing suggests that the influence of information on people’s attitude is mediated by the valence of their thoughts about the attitude object. We therefore expect that fit has a moderated indirect effect on brand attitude through thought valence. In this fit has a positive indirect influence when involvement is low (H1), but this influence is corrected when involvement is high (H2). The data for our study was collected in a laboratory experiment. We used a 2 (low vs. high involvement) x 2 (low vs. high fit) between-subjects design. Participants were 95 undergraduate students. Level of fit was manipulated by relating one brand to two causes, perceived to have high or low fit. Based on a pilot study, we selected the bicycle brand Velocidad and the causes Protection of Endangered Species Foundation (low fit) and the Traffic Safety Foundation (high fit). These causes did not differ on liking and involvement with the cause and showed the most articulated difference on fit with the brand. Involvement was manipulated by asking participants to evaluate the brand. Participants in the high involvement condition were told that this would be helpful because their own university considered cooperating with the brand, in the low involvement condition this request came from another university. This cooperation would involve discounts for students on the purchase of a bicycle. A series of t-tests showed that both manipulation of fit (t(79.48) = -8.30; p < .001; Mlow = 3.69; SDlow = 1.21; Mhigh = 5.58; SDhigh = 0.87) and involvement (t(82) = -2.73; p < .01; Mlow = 4.10; SDlow = 1.16; Mhigh = 4.77; SDhigh = 1.08) were successful. To test our predictions, we used Hayes’ (2012) PROCESS macro for SPSS. First, PROCESS conducted two linear regression analyses. The first analysis included fit, involvement, and their interaction term as independent variables. All had a significant influence on thought valence (see Table 1). The second analysis added thought valence as an additional independent variable and examined the influence of these variables on brand attitude. Only thought valence had a significant influence (Table 1). The combined results of both analyses suggested that the interaction between fit and involvement had an effect on the brand attitude, which was mediated by thought valence. Additional bootstrap analyses showed that in the low involvement condition brand/cause fit had a positive indirect effect on brand attitude (point estimate = .31; SE = .14; C.I. = .09 - .64). For the high involvement condition fit had a negative indirect effect on brand attitude (point estimate = -.10; SE = .11; C.I. = -.35 - .07). We conclude that brand/cause fit has a positive indirect effect when people are not involved, confirming H1. Involved people correct for this bias, resulting in a smaller effect of fit, confirming H2.
A NEGATION BIAS IN WORD OF MOUTH: HOW NEGATIONS REVEAL AND MAINTAIN EXPECTATIONS ABOUT BRANDS AND PRODUCTS - PEETER W.J. VERLEGH, CAMIEL J. BEUKEBOOM AND CHRISTIAN F. BURGERS Consumers like to talk about their product experiences with others. This word of mouth (WOM) can have a strong impact on product evaluations and sales (e.g., Chevalier & Mayzlin 2006), both 179
180
ETMAAL 2013
BOOK OF ABSTRACTS
practitioners and academics have great interest in understanding its effects and underlying mechanisms. Although research on WOM has a long history, dating back to the seminal work of Katz and Lazarsfeld (1955), academics have only recently begun to examine language use (Moore 2012; Schellekens, Verlegh & Smidts, 2010). Findings suggest that consumers implicitly reveal much information “in between the lines” of their WOM messages. Such formulation differences may for instance reveal speakers’ a priori brand expectations when they describe a product experience. Of particular interest is the use of negations. The Negation Bias (Beukeboom, Finkenauer & Wigboldus, 2010) demonstrates that negations usage (e.g., not stupid, rather than smart) is more pronounced when (stereotypic) expectancies are violated, compared to when expectancies are confirmed For instance, garbage men performing poorly on IQ-tests is stereotypically expected and likely described as “stupid”, for professors this would be unexpected and described as “not smart”. Importantly, negations lead message recipients to infer that the described experience was an exception, caused by situational circumstances (Beukeboom et al., 2010). In two experiments, we examined the existence of a negation bias in WOM. We expect that negations are used more often in descriptions of expectancy inconsistent than consistent product experiences. We also expect that receivers can “decode” these implicit communications: a speaker’s use of negations leads the receiver to infer that the described experience is an exception to the rule. The use of negations in inconsistent situations will dampen the impact of WOM on receivers, serving to maintain a brand’s positive or negative reputation, even in the face of WOM disconfirming that reputation. Experiment 1 examined whether people are more likely to use negations when they describe product experiences that are inconsistent (versus consistent) with their a priori brand expectations. Participants were presented with WOM vignettes that varied according to a mixed 2 (positive/negative brand expectation) x 2 (positive/negative experience) x 5 (vignettes) design. To manipulate expectations, we used pretested brands scoring either high or low on a focal attribute. For example, a customer received slow (versus fast) service in a Hilton (versus Ibis) Hotel. We created five vignettes for each condition. Each participant saw all five different vignettes within one condition. This was analyzed as a within-subjects factor. Following Beukeboom et al. (2010), participants rated how well each experience was described by a negation or an affirmation (e.g., “the service was not fast/ slow”). After rating all five experiences, they were again presented on a separate page, and participants rated the expectedness of each experience. A mixed model ANOVA revealed the predicted expectation x valence interaction, for both dependent variables (F(1,115)>9; p<.01): for positive experiences, negations were rated more applicable when participants held low (versus high) expectations, and for negative experiences, negations were rated less applicable when participants held low expectations. Effects were mediated (Hayes, 2012) by perceived expectedness: the less expected an experience, the more applicable participants rated the negations, and the less applicable they rated affirmations. Experiment 2 examined whether a speaker’s use of negations indeed leads receivers to infer that the experience was unexpected. Participants (N=123) read a scenario in which a friend commented on 180
181
ETMAAL 2013
BOOK OF ABSTRACTS
his suitcase. This comment was varied in a 2 (valence: positive/negative) x 2 (description: negation/affirmation) between-subjects design. We measured participants’ inferences about (1) the expectedness of the experience, (2) the speaker’s surprise with the experience (3) the speaker’s expectations before the experience, and (4) speaker evaluation of the experience. As predicted, participants inferred that the experience was less expected and more surprising when the speaker used negations versus affirmations (both F(1,119)> 4, p<.05). Interactions with valence were non-significant. For inferences about the speaker’s a priori expectations we do predict an interaction: negations in descriptions of negative experiences (“not fast”) signal a positive a priori expectation, but the opposite is true for positive experiences (“not slow”). Indeed, we found a (marginally) significant interaction (F(1,119)=3.55, p=.06): the effect was significant (p<.05) for negative experiences, and directionally consistent for positive experiences. Importantly, we found no significant effects of negations on inferences about the speaker’s evaluations. After all, negations do not communicate whether an experience was positive or negative. They do, however, communicate whether it was (un)expected, and whether it confirmed the speaker’s expectations. In this manner, negations provide a subtle mechanism for communicating and maintaining brand reputations.
HOW OFTEN DO YOU REALLY WATCH? COMPARING SELF-REPORTED TV EXPOSURE TO PEOPLEMETER DATA - ANKE WONNEBERGER AND MARIANA IRAZOQUI This study contributes to the discussion on the validity of self-reported measures of media use by comparing self-reported survey measures to observed measures of TV exposure stemming from electronic registration through people-meters. With average viewing durations of more than three hours per person per day, television still is one of the most important media in many Western countries – from an economical as well as from an audience perspective. While measures of TV exposure are crucial for many communication studies, there is a lack of research assessing the validity of self-reported viewing behavior. We employ a unique research design that allows to directly compare self-reported TV exposure to people-meter data based on a single sample and measured in the same time period. Previous research on the validity of self-reported measures of media use has dealt with inaccuracies of self-reports stemming from incorrect recall, estimation, or reporting of media use behavior. Studies in this area have focused on question wording, relationships between measures of different dimensions or modalities of media use, and types of survey methods such as face-to-face or web surveys. A manner to gain more accurate knowledge about the nature and the scope response errors in survey measures, is a comparison of self-reports to observational measures of media use. To date, direct comparisons of observational measures to self-reports to assess possible biases of self-reports are scarce and have not been conducted for general TV consumption. The goal of this study is threefold: First, the level of convergence validity and, therewith, the degree of overlap between survey data and observational data is assessed. To what extent do people over- or underreport their viewing behavior? Second, we test the occurrence of recency effects on self-reports. Oftentimes researchers are interested in average viewing behavior as opposed to exposure on specific days. This is why the congruency of self-reports with behavior over a longer versus a more recent reference 181
182
ETMAAL 2013
BOOK OF ABSTRACTS
period is crucial. The third aim is to detect to what extent respondents may vary systematically in their response behavior. Such systematic differences mays bias any conclusions regarding TV use and effects drawn from measures based on claimed behavior. On the one hand, levels of response might vary between sociodemographic groups. On the other hand, flaws of self-reports might stem from aspects directly related to watching TV. We use Dutch people-meter data on the individual level and the corresponding TV use variables from a survey conducted among the same respondents of the people-meter panel. The registration of individual viewing behavior with people-meters has the advantage of being to a large extent unobtrusive, independent from recall or estimation errors and, thus, yielding highly precise information on media use. Moreover, two time periods are used to test possible recency effects. Average viewing behavior over a period of eleven weeks is contrasted to viewing behavior during the week of the survey field work. Comparing the self-reported to the observed measures of TV exposure reveals that respondents typically overreport the number of days they watch TV in one week while they underreport the time they spend watching. Overreporting viewing days could to some extent be explained by a recency effect. Differences between self-reports and observed measures of the time one usually spends watching TV are systematically related to age, education, and income. Also contextual factors that shape viewing situations, such as other household members and the number of available TV sets, influenced divergences between self-reports and observed behavior. Finally, viewing motivation and behavior itself present important sources for inaccurate response behavior. Respondents with a strong interest in watching TV – as measured through their declared genre preferences – tend to overreport their viewing behavior. The more time viewers actually spend watching – based on people-meter measures – the more they tend to underreport their exposure. In short, the findings assess the level of over- and underreporting of TV viewing behavior and shed light on factors that interfere with the accuracy of self-reported viewing behavior. Inaccuracies of claimed viewing behavior related to sociodemographics but also behavioral and attitudinal aspects of viewing behavior can be considered a particular challenge for the validity of self-reported exposure measures that are frequently employed in survey designs. We suggest that these factors should be kept in mind when designing studies. Further implications for survey research on media use behavior will be discussed.
182
183
ETMAAL 2013
BOOK OF ABSTRACTS
Session #2 Thursday 7 February 15:15 – 16:15 hrs. Chair: Sonja Utz Location: Rochester (M2-10) English session
BUTT-NAKED ADVERTISEMENTS: PROMOTING TABOO PRODUCTS BY MEANS OF NUDITY - JISKA EELEN, MARIEKE L. FRANSEN AND MIKE FRIEDMAN In our research, we explore how products that are taboo and create resistance should be promoted. In two lab studies we find that taboo advertisements can benefit from using explicit nudity. In line with terror management theory, we suggest that advertisements about bodily taboos can trigger existential anxiety in consumers, which may negatively affect the evaluation of the ad. However, if nudity is present in advertising, the taboo product can actually serve as a means to cope with the experienced anxiety. Nudity in a taboo advertisement led to reduced discomfort or disgust and greater relief or happiness, and also led to more positive attitudes towards the advertisement.
"ELEVATION!" (ONLY IF YOU CARE): EXAMINING THE DETERMINANTS OF USERS' ELEVATION RESPONSES TO SHORT FILM CLIPS - ALLISON EDEN AND TILO HARTMANN Viewing exemplary moral behavior elicits a specific emotion known as “elevation” or the sensation of being ‘lifted up’ to a higher moral plane (Haidt, 2003). Experiencing elevation may persuade people to adopt intentions to act in a morally exemplary way (Algoe & Haidt, 2009; Aquino, McFerran, & Laven, 2011). Therefore, elevation through films may be an especially attractive avenue for persuasive communication aiming to increase pro-social behavior in viewers. Multiple studies have shown, however, that moral issues may elicit different responses based on the content domain of the behavior, and the salience of the particular moral concern to viewers (Haidt & Joseph, 2009; Graham, Nosek, Haidt, Iyer, Koleva, & Ditto, 2011). No study to date examines the domain of the behavior in determining elevation and subsequent behavior. The current study therefore presents two experiments designed to test the role of viewing moral excellence in particular moral domains in promoting elevation and subsequent pro-social intentions, as well as the moderating effect of the domain salience on this relationship. In study 1, we showed participants two clips illustrating moral excellence – one in the “care” domain (a young woman caring for her wartime village), and one in the “purity” domain, (a young man discussing his religiously motivated virginity). Higher levels of elevation were experienced after viewing the “care” clip (M = 3.99, SD = .87) than the “purity” clip (M = 2.49, SD = .1.00, t(118) = 8.72, p < .01). Higher levels of pro-social intentions occurred after viewing the “care” clip (M = 5.35, SD = 1.28) than the “purity” clip (M = 3.81, SD = 1.26, B = .78, SD = .12, t(118) = 6.65, p < .01). Bootstrapping analyses of the proposed mediation (Hayes, 2012) showed that both the clips affected pro-social intentions indirectly via elevation. Subsequent moderated mediation analyses showed the 183
184
ETMAAL 2013
BOOK OF ABSTRACTS
conditional indirect effect of the “care” clip on pro-social intentions through elevation was stronger if salience of “care” was high, b = .78, SE = .13, CA.95 (.53, 1.08), p < .01, and weaker if salience of “care” was low, b = .48, SE = .11, CA.95 (.28, .72), p < .01. The residual direct effect of the condition on pro-social intentions was non-significant, b = .12, SE = .13, t(118) = .88, p =.37. In total, the model explained 46% of the variance in pro-social intentions. However, participants in Study 1, predominantly Dutch university students, placed much higher weight on the “care” domain (M = 4.23, SD = .72) than on the “purity” domain (M = 3.57, SD = .75, t(118) = 8.51, p < .001) overall. This is in line with past research on moral domains in this sample (e.g. Van Leeuwen and Park, 2009). Therefore, Study 2 presented individuals only with excellence in the “care” domain - in which a girl cuts her hair for her brother, who has cancer - and compared this to a control clip featuring ordinary behavior - people saying “Hello” in different languages. The “care” clip (coded as 1) did not induce stronger pro-social intentions than the control clip (coded as 0), b = -.23, SE = .22, t(93) = -1.02, p = .31, although it was a significant predictor of elevation, b = 1.60, SE = .22, t(92) = 7.15, p < .05, and did predict pro-social behavior indirectly through elevation. The link between viewing morally excellent behavior in the “care” domain and elevation was again significantly moderated by the salience of “care” (b = 1.32, SE = .44, t = 2.96, p < .01, R2chg .05, p < .01). The effect of the clip on pro-social intentions through elevation was stronger if salience of “care” was high, b = .69, SE = .28, CA.95 (.18, 1.28), and weaker if salience of “care” was low, b = .24, SE = .15, CA.95 (.05, .63). The residual direct effect of the “care” condition on pro-social intentions remained significant, b = -.81, SE = .27, t(87) = -2.94, p < .01. In total, the model explained 11% of the variance in pro-social intentions. Together, these studies add to the literature on using media persuasively to induce pro-social behavior. We demonstrate that elevation through witnessing moral excellence contributes to subsequent pro-social intentions, and that the salience of the domain for individuals intensifies this experience. This has important implications for campaigns based on presenting excellent behavior as a method of persuading viewers to adopt pro-social intentions.
INCREASING CIVIC PARTICIPATION USING 140 CHARACTERS? AN ONLINE EXPERIMENT ON THE EFFECTS OF DIFFERENT COMMUNICATION STRATEGIES AND THE USE OF HASHTAGS - MIRJAM LASTHUIZEN, SONJA UTZ AND SANNE GAASTRA Social media enable new forms of civic participation. Citizens can give their opinions on blogs, discussion forums or Twitter and institutions like the police can engage in direct dialogue with the public. The present paper focuses on the question how the police can use Twitter to increase civic participation. We focus on two factors: the content of the tweet and the use of hashtags. With the first factor, we examine whether explicitly naming a benefit increases the willingness for participation. We differentiated between individual and group benefits. Helping the police in solving crime cases can be considered as a social dilemma. Within the social dilemma literature, a differentiation between motivational, strategic and structural solutions is made (Kollock, 1998). A motivational solution is to increase the feeling of group identity to influence cooperation (group benefit). Strategic solutions 184
185
ETMAAL 2013
BOOK OF ABSTRACTS
offer incentives for cooperation such as gaining reputation on the police website (individual benefit). The question is which of these benefits is more successful in increasing participation. According to Whiteley and Seyd (in Pattie et al, 2003): “People who feel attached to a group should be more likely to act on its behalf than those who not share that attachment”. So first we expect that citizens who engage for their community will participate more. Moreover, we expect a moderating effect of civic engagement. People who already engage do so because they value the collective good, they can be compared with prosocial individuals. Explicity naming the group benefit should further increase their participation. People who do not engage for their community can be compared to proselfs. They respond better to strategic solutions such as gaining reputation. On twitter, hashtags can be used to facilitate searching and creating discussion groups for specific topics. According to Bruns & Burgess (2011) hashtags are used to find tweets about specific issues and twitter users relate with others with the same interest. Hashtags activate a group feeling and therefore should increase participation in case of highly engaged citizens. Method An online experiment with a 3 (benefits: none, individual, group) x 2 (hashtag: yes vs. no) design was conducted. The tweet of a policeman "More burglaries in your neighborhood: Think along on bit.ly/rl46mHa" was complemented with "become the Sherlock Holmes of your neighborhood" in the individual benefit condition and in the group benefit condition with "help to increase the safety in your neighborhood”. In the hashtag condition #neighborhoodandpolice was added. Online participation was measured on three levels: retweeting the tweet, clicking the link to the website and contributing information on the website. For analysis the average civic engagement (measured with six statements) was divided in low and high engagement. Results A 3 (benefits) x 2 (hashtag) x 2 (civic engagement) ANOVA with “retweeting” as dependent variable showed main effects of benefits and civic engagement. The group benefit lead to more participation than the individual benefit. The no benefit condition did not differ from the individual or group benefit. Highly engaged citizens retweet more than lowly engaged citizens, this is consistent with our expectation that highly engaged citizens would participate more than lowly engaged citizens. ANOVA with “clicking the link” as dependent variable had the same effect of civic engagement. For “contributing information” ANOVA showed an interaction effect between benefits and civic engagement. In case of none and the individual benefit highly engaged citizens contribute more than lowly engaged citizens, but for the group benefit there was an opposite result: naming a benefit reduces contributing information when citizens are highly engaged compared to lowly engaged. Results show that online civic participation can be influenced by the content of the tweet. A group benefit leads to more retweeting and contributing information on the website than an individual benefit. For highly engaged citizens a group benefit decreases contributing information, a possible explanation is motivational crowding. Theoretical and practical implications will be presented at the conference. 185
186
ETMAAL 2013
BOOK OF ABSTRACTS
HOW INFLUENTIAL CONSUMERS CREATE BRAND-RELATED USER-GENERATED CONTENT: THE INTERPLAY BETWEEN PERSONALITY CHARACTERISTICS AND MOTIVATIONS - DAAN G. MUNTINGA, MARJOLEIN MOORMAN, EDITH G. SMIT AND PEETER W.J. VERLEGH Social media platforms such as blogs, social networks, video sharing sites, and online communities facilitate a wide variety of ways for consumers to create user-generated content (UGC). Important to marketers, the creation of brand-related UGC has been shown to shape brand perceptions (Christodoulides et al., 2011) and purchase behavior (Dhar & Chang, 2009). While it has been suggested that different types of consumers have different motivations to en-gage with brands on social media (Chen et al., 2012), there has been little research into which personality characteristics are important in this regard, and how they influence the creation of UGC. This research addresses these questions, and examines the notion that (a) consumers’ per-sonality characteristics (b) influence their motivations, which in turn (c) influence their creations of brand-related UGC. To study the role of personality and motivations in the creation of UGC, we employ the user-centric research perspective of Uses and Gratifications (U&G). U&G is the predominant theoretical framework for studying how and why individuals use media (McQuail, 2010). Moti-vation (why) is a central factor in this framework. Within social media marketing contexts, moti-vation has also been regarded as “a key element to be understood by firms willing to engage with consumers” (Kietzmann et al., 2012, p. 115). They have been shown to predict behaviors such as the creation of brand parodies (Campbell et al., 2011) and engagement in UGC (Christodoulides et al., 2012). In a recent study, Muntinga et al. (2011) show that the creation of brand-related UGC is driven by five different motivations, namely information (surveillance; opinion seeking; risk reduction), entertainment (enjoyment; relaxation; escapism) integration and social interaction (social connection; belonging; helping others), personal identity (self-expression; self-presentation; self-assurance), and empowerment (influencing others). McQuail’s U&G framework also assumes that people’s motivations for using media are shaped by their particular characteristics (who). Identifying the consumers who create UGC is critical to a successful social media strategy. After all, by creating content that is related to a brand, these users exert great influence over other consumers’ attitudes and behaviors. Based on analyses of the roles of different types of consumers in social networks (Watts & Dodds, 2007), this study uses three personality constructs related to three types of influential consumers: Ma-venism (the need to spread information to others), connectivity (the strength of one’s ties with others), and persuasiveness (one’s ability to convince/persuade others) (Boster et al., 2011). A survey was distributed among consumers that engage in UGC creation on the social networking site Hyves. 2,495 respondents (13- 68 years old, Mage = 25, 52% female) completed the survey. Measures of mavenism, connectivity, and persuasiveness (the IV’s) were taken from Boster et al. (2011). Measures for motivations, the proposed mediators, were based on Muntinga et al. (2011). Scales for information (3 items), entertainment (2 items), integration and social in-teraction (4 items), personal identity (3 items), and empowerment (3 items) were all reliable and unidimensional. The DV brand-related UGC creation was measured with a single-item: “On Hyves, I create weblogs, reviews, videos, music, pictures, and/or articles about [brand]” (1 = never, 7 = very frequently). 186
187
ETMAAL 2013
BOOK OF ABSTRACTS
A bootstrapping approach (Hayes, 2012) was used to assess specific mediating relation-ships for multiple mediators. Thus, we examined the indirect effects of the three personality characteristics on UGC creation through the five motivations. Results indicate significant total effects on UGC creation of mavenism (β = .15, SE = .04, t = 3.75, p < .01) and connectivity (β = .25, SE = .04, t = 7.11, p < .01), but not persuasiveness. Mavenism is related to motives of information (β = .46, p < .01), integration and social interac-tion (β = .46, p < .01), and empowerment (β = .32, p < .01). Connectors are driven by personal identity (β = .44, p < .01), empowerment (β = .36, p < .01), and integration and social interaction (β = .28, p < .01). The present study shows that UGC creation results from an interplay of personality char-acteristics and motivations rather than from personality characteristics (e.g., Correa et al., 2010) and motivations (e.g., Muntinga et al., 2011) working in isolation. The findings provide brand managers with useful knowledge about which types of consumers create UGC, and how these types of consumers differ with regard to their motivations for creating UGC.
Session #3 Thursday 7 February 16:45 – 17:45 hrs. Chair: Guda van Noort Location: Rochester (M2-10) English session
DISENTANGLING THE ROLE OF BRAND-TO-USER VS. USER-TO-SYSTEM INTERACTIVITY ON FACEBOOK FANPAGES - SONJA UTZ AND MARIETTE VAN TUYL Research topic: Effects of brand-to-user vs. user-to-system interactivity on Facebook fanpages Academic field: marketing communication More and more companies use Facebook fanpages in their marketing communication. Interactivity is often regarded as the key feature triggering positive responses to online communication. However, relatively little is known about the underlying processes. Van Noort, Voorveld, and van Reijmersdaal (2012) proposed online flow as mediating mechanism, but did not differentiate between user-tosystem and user-to-user interactivity. User-to-system interactivity is given when clicks result in specific actions (e.g. games, apps). User-to-user interactivity refers to communication among users, e.g. leaving comments. We argue that the effects of these two types of interactivity are driven by different underlying processes. Based on earlier research in the domain of political communication (Warnick et al., 2005), we assumed that both types of interactivity on Facebook fanpages lead to more positive reactions. H1: User-to-system interactivity results in a more positive evaluation of the Facebook fanpage and the brand, a higher intention to engage on the brand fanpage and a higher buying intention. H2: Brand-to-user interactivity results in a more positive evaluation of the Facebook Fanpage and the 187
188
ETMAAL 2013
BOOK OF ABSTRACTS
brand, a higher intention to engage on the brand fanpage and a higher buying intention Underlying processes Van Noort et al. (2012) showed that flow mediates the effects of interactivity. Flow has often been investigated in games or sports (e.g.,Cowley, Charles, Black, & Hickey, 2008), that is, situations in which people experience that they can exercise control. Therefore, we assume that flow primarily mediates the effects of user-to-system interactivity. H3: The effects of user-to-system interactivity are mediated by flow. However, flow should be less relevant when reading conversations between brand and fans. By reacting on the questions and complaints of fans, a company signals that it is willing to listen to its customers and that it cares about them. In the domain of organizational blogs, communicated relational commitment and human voice have been identified as mediating variables (Kelleher, 2009). Our last hypothesis is therefore: H4: The effects of brand-to-user interactivity are mediated by communicated commitment and human voice. Method An online experiment with a 2 (user-to-system interactivity: low vs. high) x 2 (brand-to-user interactivity: low vs. high) – design (n = 315) was conducted. Respondents viewed a movie of a person browsing the Facebook fanpage of the Bijenkorf. In the high user-to-system condition, the person used the Valentine app. In the low user-to-system condition, no app was shown. Only in the high brand-to-user interactivity condition, the brand reacted on complaints and remarks of the fans. Attitude towards the Facebook fanpage and the brand, liking intentions, participation intentions, and buying intention were the dependent variables. Human voice, communicated relational commitment and flow were assessed as potential mediators. Results A 2 (brand-to-user interactivity) x 2 (user-to-system interactivity) MANOVA with attitude towards the brand page and the brand, liking intention, participation intention, and buying intention as dependent variables and prior attitude towards the warehouse as a covariate revealed only a significant main effect for brand-to-user interactivity, confirming H2. Univariate tests showed that attitude towards the brand (Ms = 4.82 vs. 4.14), liking intention (Ms = 2.85 vs. 2.45), and participation intention (Ms = 3.32 vs. 2.69) were significantly higher in high interactivity condition than in the low interactivity condition. A higher brand-to-user interactivity resulted also in higher communicated relational commitment (Ms = 5.36 vs. 4.46), human voice (Ms =4.94 vs. 3.83), and, unexpectedly, higher flow (Ms = 2.88 vs. 2.55). Bootstrapping analyses showed that the effects on attitude towards the brand were mediated by all three mediators, the ones on liking intention by communicated commitment and flow, and the ones on participation intention by human voice and flow, thereby confirming H4. This study enhances the understanding of the role of different forms of interactivity and the underlying processes. It also provides practical implications for brand managers.
188
189
ETMAAL 2013
BOOK OF ABSTRACTS
GETTING TOO PERSONAL? ADOLESCENTS’ RESPONSES TO PERSONALIZED SOCIAL NETWORK SITE ADVERTISING - MICHEL WALRAVE, KAROLIEN POELS, MARJOLIJN ANTHEUNIS, GUDA VAN NOORT AND EVERT VAN DEN BROECK Introduction: The increasing personalization of SNS advertising has stirred up academic and policy debates about possible consequences for users’ privacy. More particularly, as adolescents may be less aware than adults of SNS commercial purposes and more freely disclose personal data for marketing purposes (Neeley, 2007; Turow & Nir, 2000; Youn, 2009), the current debate centers around the use of minors’ personal data for interactive advertising purposes and how adolescents respond to personalized advertising. In line with prior research and theory, positive effects of personalization in advertising can be assumed, because research has demonstrated that consumers have a higher tolerance of personalized messages as they are perceived as relevant (e.g., Cho & Cheon 2003). However, research has demonstrated that the tolerance of personalized communications might be tempered by privacy concerns (e.g., John et al., 2011). Therefore, the use of personal data may actually foster privacy concerns and, in turn, might boomerang persuasion effects (Goldfarb & Tucker, 2011; Phelps et al., 2001). This study aims to investigate how adolescents respond to personalized SNS advertising. Based on Altman’s privacy regulation theory (1976) we expect that there will be an ‘optimal’ level of personalization. Therefore, we hypothesize that (H1) SNS advertising will be more effective with a medium level of personalization, compared to a low and high personalization level. Furthermore, (H2) adolescents’ level of privacy concern will temper the impact of personalization on advertising effectiveness. Method: First, a pretest was conducted among 98 youngsters (aged 14-18) (1) assessing youngsters’ attitude towards the use of specific categories of personal data (e.g. forename, relation status, profile picture) for marketing purposes on SNS and (2) checking whether the brand (Coca Cola Plus) that was chosen for our main study was perceived as equally relevant for all members of the target group. Based on the pretest results, three conditions were designed: a low personalization condition (using only the forename), a medium personalization condition (forename, gender, birth date, interests and profile picture) and a high personalization condition (using nine personal data, including data respondents assessed negatively for its use in a marketing context, such as their status updates and their personal address). A total of 40 participants, aged 14-18 (M=16, 20 girls), participated in our main study that followed a counterbalanced within-subjects experimental design. Participants were invited to join a Facebook group that was set up for the study. As such, the participants’ personal data included in their profile could be accessed. Based on their data, three advertising messages were created per participant, corresponding with the three levels of personalization. The messages were shown through a video representing a Facebookpage of Coca Cola Plus. Participants were instructed to imagine that they 189
190
ETMAAL 2013
BOOK OF ABSTRACTS
encountered the advertising messages on their own Facebook page. After each message participants were asked to fill in a brief questionnaire assessing, amongst others, advertising effectiveness such as brand engagement (Zaichkowsky, 1985), attitude towards the ad (Horner in Bruner, 2005) and intention to forward the ad (Huang, Chen & Wang, 2012). Results: Repeated measures ANOVA’s show significant differences in brand engagement, attitude towards the ad and intention to forward the ad between the three conditions (all ps < .039). Bonferroni post hoc tests reveal a significant difference for brand engagement between the first (low personalization) and third (high personalization) condition (p=.031). Similarly, the attitude towards the ad was significantly more positive in the third condition (p=.023) compared to the first condition. The same holds for the intention to forward the ad (p = .014). For none of these variables differences were found based on participants’ level of privacy concern. Conclusion: The present study’s results are in line with previous research that observed a positive relationship between personalization and brand engagement (e.g. Antheunis & van Noort, 2011; Chu, 2011). Although the present study hypothesized an optimal level for the medium personalization condition, advertising effectiveness was highest in the high personalization condition. Higher personalization may thus positively influence the relevance of the campaign for the receiver (Antheunis & van Noort, 2012). Adolescents may feel entertained by the novelty of the ad that includes their personal picture and references to several aspects of their identity in a commercial of a renowned brand. Furthermore, no significant differences were found based on privacy concern. This could further endorse the ‘privacy paradox’ (Barnes, 2006; Norberg, Horne & Horne, 2007): although consumers declare to be concerned about their (online) privacy, their concern contrasts with disclosure behavior in concrete situations. In sum, the immediate benefit they experience by the personalized ad may overshadow general privacy concerns.
IN-GAME ADVERTISING IN CONTEXT: THE IMPACT OF SOCIAL SETTING AND BRAND TYPE ON THE EFFECTIVENESS OF IN-GAME ADVERTISING - LAURA HERREWIJN AND KAROLIEN POELS Digital gaming has become one of the largest entertainment sectors worldwide. The ever-increasing popularity of games, in combination with their long shelf-life and fun, interactive and immersive nature, has made the medium into an interesting vehicle for advertisers[1,11,12,14,17,18]. Consequently, the practice of in-game advertising (IGA) is predicted to grow steadily over the next several years. However, IGA revenues are still relatively low compared to advertising revenues spent on other media, implicating that digital games are still not fully explored as an advertising vehicle[3,5,16]. Additionally, academic research on IGA seems to suggest that there is no such thing as an optimal, “one-fits-all” formula for the successful integration of IGA. IGA effectiveness often depends on several factors related to the advertising context, such as the type of brand or ad that is integrated, the genre of the game, the setting in which the ad is encountered, etcetera [7,10,11,12,14,15,18]. The aim of the study is to further analyze IGA effectiveness by addressing two of these contextrelated factors. First of all, we look at the effect of social setting. Traditional advertising research has demonstrated that the presence of other people alters the impact of advertising[2,6,8,9,13], and that this effect can either be negative (distraction model)[2] or positive (social influence model)[6,9]. A 190
191
ETMAAL 2013
BOOK OF ABSTRACTS
key concept that we take into account is shared involvement, or the involvement that is derived from a player’s awareness of and interaction with others in a game[4]. Further, we also test the impact of interactivity (play versus watch), since research suggests that this also influences how people process IGA[16]. Second, we want to check which brands benefit most from IGA (in a social setting). Several studies have demonstrated that IGA effectiveness is highly related to the type of brand that is integrated into the game. This effectiveness can differ according to the (inter)national nature of brands, their personal relevance, familiarity, etcetera[11,12,14,16]. In the current study, we will examine two factors related to brand type: familiarity and social desirability. In order to analyze the effects of these factors on IGA effectiveness, we manipulated the social setting in which participants played the game Little Big Planet 2 in a between-subjects experiment (N = 91), resulting in four conditions: people either played the experimental game alone (single play condition); together with one other person (multi play condition); or alone while another person watched (public play: player and public play: observer conditions). When people participated in pairs, they were acquainted with each other and were of the same gender. Additionally, we included brand type and product category as within-subjects factors. The experimental game contained four brands from two different product categories (neutral versus social desirability brands) with varying degrees of familiarity. After finishing the experimental level, participants were asked about their shared involvement while playing the game, and their awareness (i.e. recall, recognition) of and attitudes towards the integrated brands. Our results show that the manipulation of social setting significantly affects shared involvement on the one hand, and brand awareness on the other. This effect is particularly situated between the player versus observer conditions: people who merely observed the game play of another person experienced less shared involvement than people in the player conditions (the single play condition excluded), and were able to recall and recognize more brands. Subsequent mediation analyses showed, however, that this was primarily due to the different roles of players versus observers, rather than to an effect of shared involvement. An explanation for this might be that the people in the observer condition did not actively control or interact with the game directly. The cognitive resources that are required for active control of a digital game may inhibit the processing of IGA: players spend more time controlling the game and have less time for attending to ads than observers[16]. These results seem to suggest that watching (while others play) a game has a positive impact on brand awareness. This has interesting implications for the practice of IGA. First of all, players often pause their game play and take turns playing and watching games with their friends, giving advertisers the advantage of reaching both passive and active audiences. Furthermore, digital games often feature non-interactive cut scenes, which may offer opportunities for the integration of brands[16]. Finally, we found that IGA has a different effect when different brand types are integrated, although only with respect to brand familiarity. Our findings show that familiar (versus less familiar) brands
191
192
ETMAAL 2013
BOOK OF ABSTRACTS
generate a better brand awareness. This can be explained by the fact that well-known brands are accessible objects that automatically attract attention, making them easier to remember[11,12,16].
THE MYTH OF E-MAIL PERSONALIZATION AND INTERACTIVITY - REMCO DE VRIES, TOM BAKKER, DAMIAN TRILLING AND PIET BAKKER Companies use e-mail on a massive scale to communicate with customers (EmailStatCenter, 2011). It is fast, cheap and simple; thousands or even millions of customers can be reached almost immediately. Although e-mail is an efficient way to reach large audiences, the effectiveness of this way of communication is not yet optimal: Many people do not respond in the desired way to the message. To increase the effectiveness of corporate communication, personalization and interactive features are often used (Lee, 2005; Liu & Shrum, 2002; Maslowska, Smit, & Van den Putte, 2011). Addressing customers for instance with their real name instead of “dear customer” and offering interactive options from which customers can choose, are thought to be appealing to users and increase the chance that e-mails are opened, read and acted upon (Postma & Brokke, 2002). Previous research on these features, however, is inconclusive. In some cases, personalization and interactivity do seem to result in a more positive attitude of the readers towards the message and the company (Kruikemeier, Van Noort, Vliegenthart, & De Vreese, 2012; McMillan & Hwang, 2002; Sicilia et al., 2005), but in other cases no significant relations were found (Coyle & Thorson, 2001; Fortin & Dholakia, 2005). This study hypothesizes that these mixed results are caused by shortcomings of earlier research that was predominantly conducted within controlled laboratory environments with mostly students as respondents. First of all, these groups differ significantly from ‘real’ customers and the conditions under which e-mails are received, opened (or not), read and acted upon also differ very much from real-life conditions. Secondly, e-mails are, because of strict legislation, mostly received by existing customers, meaning that receivers already have a relation with the source. This condition cannot be replicated in a laboratory setting. Furthermore, some of the research dates from the beginning of this century. Customers may have developed different attitudes toward privacy, interactive features, or handling mail in general over the years. This also could affect outcomes. To overcome these shortcomings, a unique real-life experiment was conducted, using the database of a Dutch telecomoperator and sending out four different versions (with and without personalization and interactive features) of their regular newsletter. One day after the newsletter, all customers received a survey in which their attitudes towards the company, the product and the newsletter were measured. The results show – after analyzing data of more than 1200 completed surveys – that neither personalization nor interactivity did have any significant effect on the attitude of the receivers. We explain these results by pointing at the difference between laboratory experiments and real-life experiments, particularly the differences between the respondents and differences between conditions under which the e-mails are handled. In experiments respondents are asked to open and read e-mails sitting behind a screen in a controlled environment and respond to them immediately. In real life, more than half of the respondents do not even bother to open e-mails. Also, we found that people do not respond immediately to the message – in fact some respondents waited more than two weeks before completing the survey. Messages – when opened – are not always read from 192
193
ETMAAL 2013
BOOK OF ABSTRACTS
the big screen of a desktop or laptop computer, almost 40% was read from mobile phones or tablets. Also the pre-existing attitude towards the company that is sending the mail could be a determining factor. For customers who are concerned about privacy, personalization could even be a negative factor. Customers probably receive many e-mail messages and want to handle them as quickly as possible, which will not make them very interested in interactive features. Laboratory experiments are very suitable to build and refine theories, but their ability to test whether these theories do indeed predict real-life behavior are limited. The practical consequences are far-reaching. Companies could save substantially on e-mail marketing by skipping personalization and making mails simpler by not using too many interactive features. Finding out whether non-personalized and simple non-interactive mails indeed yields better results, could be the topic of further research.
Session #4 Thursday 7 February 14:05 – 15:05 hrs. Chair: Eva van Reijmersdal Location: Auckland (M3-04) English session
MINDLESS RESISTANCE TO INFLUENCE: WHEN LOW SELF-CONTROL DECREASES YIELDING TO PERSUASIVE COMMUNICATION - LOES JANSSEN AND BOB M. FENNIS Living in a consumer society, we are confronted on a daily basis with innumerable forms of persuasive communication. Advertisers, sales representatives, marketers, politicians, fundraisers, and other influence professionals use an impressive array of persuasive techniques to make us buy their products, use their services and support their ideas. Although part of their messages is welcomed by its recipients, the overwhelming majority of these influence attempts can be categorized as unsolicited persuasion. To help people guard their minds and wallets, it is essential that we understand the process of resisting and yielding to these types of persuasive communication. Recent research has shown that resisting (unwanted) persuasion requires active, controlled guidance by the self, and attempts at resistance are more likely to fail when resources for self-control are low (for a review see Burkley, Anderson, & Curtis, 2011). Low self-control (Baumeister, Vohs, & Tice, 2007) has been found to increase susceptibility to diverse forms of persuasive communication, from attitude change in response to a counter-attitudinal persuasive message (Burkley, 2008; Wheeler, Briñol, & Hermann, 2007) to yielding to (unsolicited) requests to invest time, effort, or money supporting previously unknown causes (Fennis, Janssen, & Vohs, 2009; Janssen, Fennis, Pruyn, & Vohs, 2008). However, the present research demonstrates that under certain conditions, low selfcontrol may facilitate, rather than hinder, resistance to persuasive communication. When taking a closer look at the work of Fennis et al. (2009) and Janssen et al. (2008), they showed that the effect of low self-control on compliance was qualified by the presence of a compliance193
194
ETMAAL 2013
BOOK OF ABSTRACTS
promoting heuristic in the influence context. For example, low self-control participants were willing to spend more of their time volunteering in lab studies than high self-controls when the experimenter previously did them a favor (manipulating the salience of the reciprocity heuristic, Fennis et al., 2009, Experiment 4). Importantly, when no such heuristics were present, low selfcontrol did not affect compliance. In other words, it is not the case that people low in self-control invariantly yield to whatever type of persuasive communication they are exposed to, and those high in self-control always evidence higher levels of resistance. Rather, both low and high self-control individuals show equal levels of compliance when the influence context does not harbor salient decisional heuristics. This suggests that even low self-controls can see through blunt, direct attempts at being influenced, however, in contrast to their high self-control counterparts, low self-control individuals are more easily swayed by salient, easy-to-process information in the decision environment that helps them guide their behavior. This makes the question imperative whether people low in self-control would similarly be prone to use decision heuristics that promote resistance to persuasion, rather than compliance. Three experiments tested the hypothesis that low self-control facilitates mindless resistance to persuasion when the influence context harbors resistance-promoting heuristics. Experiment 1 used a 2 x 3 between-subjects design. Low self-control was induced with a self-control task and participants were asked whether they would be willing to donate money to charity. This request was either preceded by the activation of a compliance-promoting heuristic (participants received an unexpected favor which activated the principle of reciprocity, cf. Gouldner, 1960), a resistance-promoting heuristic (participants were withheld from an expected favor which violated the principle of reciprocity), or no heuristic. In line with previous findings, a compliance-promoting heuristic increased the willingness to donate for those low in self-control. However, these participants were less willing to donate money to charity when the resistance-promoting heuristic was activated. In Experiments 2 and 3, instead of manipulating self-control, dispositional self-control was measured and participants were confronted with a request to volunteer in future lab studies. For half of the participants, this request was accompanied by a resistance heuristic. In Experiment 2, this was a violation of the reciprocity principle (the person that made the request already owed participants a favor), and in Experiment 3 this was a message disclaimer in the form of negative social proof (cf. Cialdini, 1993): "18% of our students volunteer in lab studies, 82% doesn’t volunteer". Participants low in dispositional self-control complied less with the request (Experiment 2) and generated more negative responses towards it (Experiment 3) when it was accompanied by a resistance heuristic. There were no differences for those high in self-control. The present research suggests that resistance to persuasion is not an invariantly effortful and consciously guided process that requires self-regulation to succeed. On the contrary, in certain conditions resistance constitutes an act of mindlessness (Langer, 1992), and is more likely when one does not carefully scrutinize the influence attempt.
194
195
ETMAAL 2013
BOOK OF ABSTRACTS
IS BEING BAD AS BAD AS BEING GUILTY? THE ROLE OF IDENTIFICATION AND JUST ENDINGS IN EVOKING EMOTIONS IN NARRATIVE PERSUASION - HANS HOEKEN There is ample evidence that narratives, whether in the form of short stories, books, television series or movies, can influence the beliefs, attitudes, intentions, and behavior of their audience (e.g., Appel & Richter, 2007; De Graaf et al., 2009, 2012). Now that the persuasive potential of narratives has been firmly established, research is aiming to identify the mechanisms responsible for these effects. In various studies, the emotions evoked by the narrative have been found to correlate strongly with the persuasive impact (Busselle & Bilandzic, 2009; De Graaf et al., 2009). The question is how narratives evoke emotions? In the studies revealing the importance of emotions, these emotions are related to identification with the character. A straightforward interpretation of these results would be that the stronger the people feel for the character, the stronger the emotions they experience. For instance, if the narrative has a happy ending for the main character, people experience more joy the more they have identified with the character. Conversely, in case of a sad ending, people may experience more sadness and anger the more they have identified with the character. However, based upon theories of media enjoyment and entertainment, a different route could be at work. Raney (2004) has developed a disposition theory of media enjoyment in which character liking and moral evaluations play a decisive role. Basically, Raney predicts that a narrative evokes positive emotions (e.g., joy) if the interaction between character type and story outcome are in accordance with our view of a just world. That is, we experience positive emotions if the good guy gets rewarded and the bad guy gets punished. Conversely, if the bad guy gets away with his bad deeds while the good guy gets punished, the audience perceives this as a mismatch with our view of a just world; as a result negative emotions (e.g., anger, sadness) are evoked. According to Raney’s disposition theory, the extent to which a story with a negative outcome for the main character evokes negative emotions depends on the characteristics of this character. Two characteristics may be important in this respect: culpability and likeability. If the character is considered to have brought about the trouble himself, the negative outcome can be regarded as in accordance with a just world and no negative emotions would be evoked. The question is whether a similar effect is to be expected if the character is unsympathetic (but cannot be held responsible for his troubles)? In two experiments, the question was addressed whether identification with the character and/or the perception of the outcome as in accordance with a just world serve as mediator variables in the evoking of emotions. In both experiments, a story was used in which the main character eventually dies because of the unavailability of a donor heart. In experiment 1 (Ss = 116), likeability was manipulated by having participants read a story in which a woman waiting for a donor heart either likes or dislikes the children from her partner’s first marriage. This manipulation was successful in influencing character likeability (p < .001), but failed to influence the identification with the character (p = .11), the perception of a just world (p = .08), and the evoking of negative emotions (p = .25) and
195
196
ETMAAL 2013
BOOK OF ABSTRACTS
emotional intensity (p = .35). In experiment 2 (Ss = 115), the character’s culpability was manipulated by either having participants read a story about a man suffering from a heart disease that resulted from an unhealthy lifestyle or from hereditary factors. In the former case, the man was considered more culpable (p < .001) and less likeable (p < .001). Furthermore, participants identified to a lesser extent with the culpable character (p < .05), the outcome was considered more just (p < .001), and evoked less negative (p < .01) and less intense emotions (p < .05). These results enable a multiple mediator analysis as propagated by Preacher and Hayes (2008). The results of this analysis reveal that identification is a significant mediator both for negative emotions and emotional intensity whereas perception of just ending is not. If narrative persuasion is, at least partly, fueled by emotions, the results of the experiments reported here reveal insights into how narratives can evoke them. The results underscore the importance of identification as a mediator in this process.
DEVELOPMENT OF A SCALE TO MEASURE STRATEGIES TO RESIST PERSUASIVE COMMUNICATION MARIEKE FRANSEN, PEETER VERLEGH AND CLAARTJE TER HOEVEN Over the years, persuasion researchers have mostly been concerned with improving persuasion attempts by investigating factors that contribute to attitude change and the mechanisms underlying such change (see Petty and Wegener, 1998 for a review). The persuasion models that have been developed so far (e.g., Chaiken, 1980; Greenwald, 1968; Kruglanski and Thompson, 1999; Petty and Cacioppo, 1984; Petty, Ostrom, and Brock, 1981) do well in explaining why attitude change occurs but they do not explain why certain attitudes are hard to change and how resistance might be overcome. Persuaders could benefit from this knowledge by developing strategies that overcome consumers’ resistance whereas consumers may use this knowledge to develop strategies that help them to resist persuasive messages more effectively. Jacks and Cameron (2003) were one of the first to systematically investigate how people resist persuasive messages that counter argue their own opinion. They asked participants to write an essay on how they would resist changing their opinion when someone challenges it. Based on the results they identified six strategies people use to resist persuasion (i.e., attitude bolstering, counter arguing, selective exposure, social validation, source derogation, and assertions of confidence). Subsequently, they developed a questionnaire assessing likelihood ratings of adopting the different strategies and tested whether the strategies predict actual resistance. However, Jacks and Cameron (2003) focused on societal issues like abortion and the death penalty; issues on which people generally have a well-defined and strong attitude. It has, however, remained unclear whether and which resistance strategies people use when exposed to other kinds of persuasive messages on other kinds of topics. Since people are exposed to, for example over 1000 advertisements each day (Kotler & Keller, 2005) it would be interesting to examine whether and which resistance strategies people use when exposed to these kind of persuasive messages. Based on the literature on resisting advertising (e.g., Ahluwahlia, 2000; Friestadt & Wright, 1994; Kirmani & Campbell, 2004), we identified four additional resistance strategies (i.e., avoidance, distraction, invoking persuasion knowledge, and message derogation) that might be relevant in resisting persuasive communication. 196
197
ETMAAL 2013
BOOK OF ABSTRACTS
The aim of the present research is to further develop and validate a scale to assess resistance strategies that people adopt in resisting persuasive messages. Based on a relevant literature search and brainstorming, we first developed new items to assess the four additional resistance strategies. Based on expert discussions, we adjusted the items used by Jacks and Cameron (2003) so they could be applied in the current context. This resulted in a total of 33 items to measure the 10 different resistance strategies. To examine these items, we conducted a survey in which we exposed participants (N = 250) to an advertisement (different advertisements were used as a between-subjects factor), and asked them how they were most likely to react when resisting the advertisement. The 33 items were presented in random order. We factor-analyzed the results and found a nine-factor solution which in total explained 72% of the variance. The nine identified factors corresponded with the resistance strategies. The resistance strategy ‘attitude bolstering’ was, however, not observed as a strategy in the current sample. We additionally found that participant’s experienced negative affect when exposed to the advertisement increases the likelihood that participants use the resistance strategies. Moreover, as expected, we observed a negative relation between involvement with the advertised product and the likelihood that participants adopt the avoidance strategy. Consistent with the literature (Hong, Giannakopoulos, Laing, & Williams, 1994), we found a negative relationship between age and the likelihood that participants adopt the different resistance strategies. Currently, further research is under way in which this scale is further developed and validated in large representative consumer samples, and correlated with personality traits such as psychological reactance, consumer self-confidence, and resistance towards persuasion.
IMPLICIT PERSUASION: HOW NODDING THE HEAD AFFECTS PEOPLE’S OWN ATTITUDES ANNEMARIE WENNEKERS There are different ways in which people could persuade themselves or others. Such persuasion may take place through a deliberative process: by generating powerful arguments for our points of view, we hope that the perceiver – after weighing these arguments – reaches the same position or attitude on the subject. However, our attitudes and behaviors are not always a result of deep thought processes. Instead, we often may not even be aware of what drives how we feel or behave. How could people be persuaded through a more implicit process? In the present research project, we addressed this question by investigating how persuasive bodily signals affect people’s implicit attitudes. A research field in which the contrast between explicit and implicit attitudes has been shown to be of extreme importance is that of prejudice. Prejudice can be defined as the negative attitudes people may have toward people from other groups (also called the out-group). People have the (natural) tendency to categorize others into groups, based on for example ethnicity, social role or gender. Unfortunately, many problematic forms of prejudice result from these categorizations. Nowadays, in most western societies the norm is to be(have) non-prejudiced. Still, people might experience negative feelings or have negative thoughts about other groups. This could result in a difference between their explicit attitudes and their more automatic associations with a group. Research has shown that explicit and implicit attitudes toward people from other groups predict 197
198
ETMAAL 2013
BOOK OF ABSTRACTS
different types of behavior. First, explicit attitudes have been shown to guide more deliberative behaviors toward people from the out-group. For example, if people explicitly state to like a group, they are likely to say more friendly things to a person belonging to that group. In contrast, the same person's implicit attitude may be more negative, and this negativity has been found to affect more spontaneous behaviors toward a person from the out-group, such as avoiding eye contact or keeping physical distance. When trying to change these more implicit negative attitudes, it may not be effective to give arguments of why prejudice is wrong or why someone should accept other groups. In fact, most people will very likely already agree with you on this point. Are these automatic negative associations then unalterable? For a while, psychologists indeed believed that implicit attitudes represented stable attitudes that were formed through a lifelong process of experience and were therefore hard to change. However, in recent years it has been shown that implicit attitudes are malleable and subject to changes in the person or the environment. For example, implicit negative associations with an out-group have been found to diminish as a result of procedures such as evaluative conditioning, in which pictures of the other group are repeatedly coupled with positive unrelated pictures. However, if implicit prejudice primarily predicts negative nonverbal behaviors, could these behaviors be used to change the underlying attitude? Findings in the area of embodied cognition have indeed shown that people’s attitudes do not only predict their behaviors; conversely people’s behaviors also play a central role in determining their attitudes. In the present research project, we took an embodied cognition approach to the reduction of implicit negative attitudes toward people from other ethnic groups. Instead of coupling positive pictures to pictures of the out-group, we used people's own persuasive bodily responses to alter the negative associations they might have with the other group. Concretely, we had Dutch participants repeatedly following a moving circle, which moved either horizontally to induce a shaking movement, or vertically to induce a nodding movement. The circle was presented after seeing a Moroccan name, a Dutch name, or a neutral word. Participants had to pronounce the name or word and then follow the circle. We repeatedly coupled either nodding or shaking movements to the Moroccan name. Before and after this head movement task, we measured people’s implicit attitudes toward Moroccans using a reaction-time based measure (Implicit Association Test; IAT). The results of two experiments showed that repeatedly nodding the head in response to Moroccan names resulted in more positive implicit associations toward this group. For people who repeatedly shook the head or performed a control task, implicit attitudes remained unaltered. In conclusion, these studies show that people may be implicitly persuaded by their own communicative bodily signals. These results will be discussed from the perspective of theories on evaluative conditioning, self-perception and embodied cognition.
198
199
ETMAAL 2013
BOOK OF ABSTRACTS
Session #5 Thursday 7 February 15:15 – 16:15 hrs. Chair: Karolien Poels Location: Auckland (M3-04) English/Dutch session
THE ROLES OF BRAND TRUST AND BRAND EQUITY IN GREEN BUYING BEHAVIOR: NATIONALLY CERTIFIED LABELS VERSUS FOOD RETAIL BRANDS - JOS BARTELS, MACHIEL REINDERS AND KAREN HOOGENDAM The use of branding as a marketing tool has received much attention during the last few decades (Aaker, 1997; Chaudhuri & Holbrook, 2001; Hoeffler & Keller, 2002; Keller, 2002; Yoo, Donthu, & Lee, 2000). More recently, branding studies have been conducted in an organic context (Bartels & Hoogendam, 2011; Larceneux, Benoit-Moreau, & Renaudin, 2012). Simultaneously, a debate in retailing and policy has emerged concerning the competitive advantages and disadvantages of food retail brands (i.e., private label brands), which have grown in power relative to premium or national brands (Ailawadi, Neslin, & Gedenk, 2001). Similarly, increased awareness of organic food retail brands is believed to endanger consumers’ trust in national labels. The present study investigates 1) the roles that brand trust and brand equity play in organic buying behavior and 2) the different impacts of these factors when comparing national certified labels and food retail brands. We conducted an international online panel survey of consumers in the Netherlands (N=508), Germany (N=519), the Unites States (N=512), Malaysia (N=403) and Singapore (N=403). Apart from questions about the respondents’ background, the questionnaire covered three topics: brand equity, brand trust and organic buying behavior. For the western countries, brand equity and brand trust seem to be higher for the nationally certified labels than for the food retail brands. However, we did not find any differences between the nationally certified labels and the food retail brands in the Asian countries. There are moderately to strongly positive correlations among brand equity, brand trust and organic buying behavior. This result implies that the more consumers experience brand equity and think that a specific brand is trustworthy, the more they will buy organic food products in general. The relationship between brand equity and buying behavior and between brand trust and buying behavior did not differ between the nationally certified labels and food retail brands. This result implies that the food retail brands do not necessarily hinder the success of the nationally certified labels. In contrast, both brand categories could equally stimulate organic food buying behavior. Brand equity and brand trust for a specific product play an important role in stimulating organic food buying behavior and that investment in brand equity and brand trust seems to pay off. More specifically, product branding can be an important strategy for retailers to increase the market share of organic food products. Therefore, marketing strategies should focus on building both brand equity and trust for nationally certified organic labels and specific food retail organic brands.
199
200
ETMAAL 2013
BOOK OF ABSTRACTS
EXPLAINING THE IMPACT OF CHARACTER-PRODUCT CONGRUENCE IN HEALTHY FOOD PROMOTING PICTURE BOOKS - SIMONE DE DROOG, MONIEK BUIJZEN AND PATTI VALKENBURG Increasing evidence indicates that brand characters can enhance young children’s food choices (e.g., Carruth, Skinner, Moran, & Coletta, 2000; Kotler, Schiffman, & Hanson, 2012; Roberto, Baik, Harris, & Brownell, 2010). However, a character’s success can depend greatly on how well it fits with the associated product, often referred to as character-product congruence (De Droog, Buijzen, & Valkenburg , 2012). Children perceive a character and a product to be most congruent, when the two form a familiar concept, such as a rabbit and a carrot (De Droog et al., 2012). Recently, a theoretical model was introduced explaining the impact of character-product congruence on young children’s healthy food behaviors (De Droog, Buijzen, Opree, & Valkenburg, 2011). This model was based on the processing fluency perspective that congruent stimuli are processed easier than incongruent stimuli, and that easy processing is inherently pleasurable (Jacoby, Kelley, & Dywan, 1989; Reber, Winkielman, & Schwarz, 1998). The model proposes that the pleasant feeling children experience when processing a congruent character-product combination manifests itself in a positive automatic response toward the product. In turn, this automatic response feeds into children’s more conscious responses, evoking a positive elaborate response toward the product that enhances children’s choice for that product. The aim of the present study was to test this character-product congruence model in a more naturalistic setting. Specifically, we investigated whether the model could explain the impact of congruent character-product combinations in picture books used in shared reading sessions at school. The fifty-two children (aged 4-6 years) participating in these sessions were assigned randomly to a picture book with either a congruent or an incongruent character-product combination. The product was a carrot, and the characters were either a rabbit (congruent) or turtle (incongruent). The book was read on five consecutive days. The conceptual character-product congruence model was tested twice with children’s affective responses and choices measured after one and five book exposures. The structural equation modeling results indicated that the model explained the impact of characterproduct congruence in picture books after one book exposure, but not after five exposures. Specifically, after one picture book exposure, character-product congruence was able to enhance children’s carrot choice via an initial positive automatic response toward the carrot. This mechanism disappeared after five exposures. We believe this was the result of children’s increased familiarization with the incongruent character-product combination as the week progressed, likewise inducing automatic affective responses comparable to the congruent combination.
“YOU CAN MAKE A DIFFERENCE!” APPLYING CUSTOMIZATION TO FUNDRAISING MESSAGES EWA MASLOWSKA, EDITH SMIT AND BAS VAN DEN PUTTE Charity organizations struggle with raising funds for their services due to cuts in governmental support and severe competition. Potential donors feel overwhelmed by the increased number of donation requests (Abdy & Barclay, 2001; Sargeant & Kähler, 1999). Therefore, charities are trying to 200
201
ETMAAL 2013
BOOK OF ABSTRACTS
find new ways of eliciting help from individual donors, and in doing so, they follow the road of commercial marketers by employing marketing tactics and new media technologies (Malafarina & Loken, 1993; Stead et al., 2006). A few recent studies conducted on the topic imply that the effectiveness of fundraising campaigns is improved by customizing communication (Hill & Moran, 2011; Verhaert & den Poel, 2011). We define customized communication as communication created for individual receivers based on the information about their characteristics and preferences (based on Kreuter, 1999). We identify two types of customization: personalization and tailoring. Personalization is based on incorporating personal cues within a message, such as name. Tailoring entails adaptation of the content of a message to personal preferences, such as beliefs. Noar, Benac and Harris (2007) showed that customized health messages are more effective than no-treatment conditions or generic messages. The effectiveness of customization has also been proved for marketing communication (e.g., Howard & Kerin, 2004). However, we do not know whether customized fundraising messages are effective. The expectation is that customizing communication is a good way to improve the effectiveness of fundraising marketing, which is the focus of this study. We compared the effectiveness of two customization strategies (personalization and tailoring) and a combination of those. Moreover, we tested the underlying mechanism of customization. Based on a study of Dijkstra (2005), in which the effect of the tailoring strategy was shown to be completely mediated by the extent to which it was perceived as being directed at the recipient personally, we tested the role of the perception that the message was personal. Customized messages are expected to be perceived as personal, because they incorporate personal information and as such are able to be seen as “for me” and to evoke the feeling of “me-ness” (Hawkins et al., 2008; Petty, Wheeler & Bizer, 1999). Based on a common assumption that customized communication is processed via the central route of the elaboration likelihood model (Tam & Ho, 2005), we further expected that the perception of the message as personal would attract attention. Increased attention was hypothesized to increase processing, which would exert persuasive effects. The study was conducted among bachelor students (N = 208, 80.3% female, Mage = 20.10, SDage = 1.54). Participants were randomly exposed to one of the following versions of a message promoting donating to The Dutch Cancer Society (KWF): generic (n = 53), personalized (n = 52), tailored (n = 53) and combined (n = 50). Next, they answered filler questions, after which they completed a survey measuring dependent variables. The messages were based on existing promotional materials of KWF and differed with respect to condition-specific features only. In the personalized condition personal cues were included, such as name, while in the tailoring condition individual beliefs about donating were addressed. These beliefs were derived from a pre-test. AMOS was used for testing the expected relations. The obtained model fitted the data well, χ2(29, N = 208) = 50.498, p = .008; RMSEA = .06 with p-close = .259; CFI = .96. The paths coefficients showed that both the personalized and combined conditions significantly increased perception of personalization when compared to the generic message. The strength of the two effects did not differ significantly from each other; they were both equally strong predictors of the mediating variable ‘perceived personalization’. The results also showed that the personalized and combined strategies indirectly led to persuasion effects. Perceived personalization 201
202
ETMAAL 2013
BOOK OF ABSTRACTS
increased the respondents’ attention and had a positive effect on attitude towards the advertisement. Furthermore, our model showed that the increased attention triggered more positive thoughts. Consequently, positive thoughts evoked a more positive attitude towards the message. Finally, the data confirmed that ad attitude positively affects intention through brand attitude. All these effects summed up to a positive and significant total effect of the personalized and combined strategies on attitude towards the message. These results suggest that customizing fundraising messages by adding personal cues makes them more persuasive, but that tailoring alone does not. This process is completely mediated by the perception of being personally addressed. Moreover, our results showed that this perception can directly lead to a more positive attitude. Consequently, companies aiming to create more effective fundraising messages should consider applying personalization.
IK HEB DUURZAME KLEDING GEKOCHT: MAG IK NU MET DE AUTO NAAR DE SUPERMARKT? MARIJN H. C. MEIJERS, MARRET K. NOORDEWIER, EDITH G. SMIT AND PEETER W. J. VERLEGH Overheden en ideële instellingen proberen mensen aan te zetten tot duurzaam gedrag: recycling, groene energie en minder autorijden zijn allemaal in campagnes voorbijgekomen. Maar wat nu als het creëren van het gewenste gedrag in het ene duurzame domein leidt tot ongewenst gedrag in een ander duurzaam domein? Recent onderzoek heeft aangetoond dat na een morele beslissing, mensen zich vaak gevrijwaard voelen vervolgens een immorele beslissing te nemen. Mensen die aangaven op Obama te stemmen, voelden zich bijvoorbeeld vervolgens gevrijwaard om een racistische beslissing te nemen (Effron, Cameron, & Monin, 2009). Deze effecten worden verklaard door het idee dat mensen door het eerste morele gedrag krediet opbouwen dat hun moraliteit aantoont, waardoor ze zich immoreel gedrag kunnen permitteren – ze hebben immers al laten zien dat ze een moreel persoon zijn (Merritt, Effron, & Monin, 2010). Duurzaam gedrag wordt vaak gezien als moreel gedrag (Schmuck & Schultz, 2002), daarom is het aannemelijk dat er binnen het domein van duurzaamheid ook een dergelijk vrijwaringseffect zou kunnen ontstaan. Zo kun je je bijvoorbeeld voorstellen dat mensen die net hun afval gescheiden hebben, vervolgens de fiets laten staan en toch maar de auto nemen naar de supermarkt: ze zijn immers al zo milieuvriendelijk bezig geweest. Dit idee wordt ondersteund door onderzoek dat aantoont dat er vrijwaringseffecten plaatsvinden tussen algemeen moreel gedrag en duurzaam gedrag. Zo leidt duurzaam gedrag tot meer liegen en zijn mensen die zich moreel goed voelen eerder geneigd zich onduurzaam te gedragen (Mazar & Zhong, 2010; Sachdeva, Iliev, & Medin, 2009). In Studie 1 onderzochten we of dit soort vrijwaringseffecten niet alleen tussen algemeen moreel gedrag en duurzaam gedrag bestaan, maar ook binnen het domein van duurzaamheid. Respondenten werden gevraagd een outfit te kiezen in een nagebouwde webwinkel met een voorjaarscollectie (controle conditie) of een voorjaarscollectie van biologisch katoen (duurzame conditie). Vervolgens vroegen we respondenten naar hun duurzame intenties. Het bleek dat respondenten die een outfit gekozen hadden in de duurzame webwinkel vervolgens minder duurzame intenties hadden dan respondenten die in de gewone kledingwebwinkel een outfit gekozen hadden. Deze studie laat zien 202
203
ETMAAL 2013
BOOK OF ABSTRACTS
dat vrijwaringseffecten ook bestaan binnen het domein van duurzaamheid: duurzaam gedrag kan leiden tot onduurzaam gedrag. Echter, vrijwaringseffecten treden niet altijd op: er is ook onderzoek dat laat zien dat duurzaam gedrag juist leidt tot meer duurzaam gedrag (Cornelissen, Dewitte, Warlop, & Yzerbyt, 2007). Dit roept de vraag op wanneer een vrijwaringseffect vooral zal optreden en wanneer het minder waarschijnlijk is. We denken dat dit afhangt van het zelfbeeld van de persoon: wanneer iemand zichzelf als een niet-duurzaam persoon ziet zal deze eerder een vrijwaringseffect vertonen dan wanneer iemand zichzelf als een duurzaam persoon ziet. Onderzoek laat namelijk zien dat wanneer duurzaamheid een belangrijk onderdeel van iemands zelfbeeld is, dit consistenter iemands gedrag aanstuurt dan wanneer duurzaamheid minder belangrijk is voor iemands zelfbeeld (Verplanken & Holland, 2002). Deze voorspelling biedt aanknopingspunten voor de opzet van effectievere campagnes, die minder aanleiding geven tot vrijwaringseffecten. We testen deze voorspelling in Studie 2. Respondenten kregen eerst een vragenlijst die hun zelfbeeld betreffende duurzaamheid mat. Hierna kregen respondenten in de duurzame conditie een advertentie van duurzame sneakers te zien en respondenten in de controle conditie een advertentie van gewone sneakers. De respondenten werden gevraagd de advertentie goed te bekijken en zich voor te stellen dat ze de sneakers hadden gekocht. Tot slot maten we hun duurzame intenties. We vonden wederom dat respondenten in de duurzame conditie minder duurzame intenties rapporteerden dan respondenten in de controle conditie. Dit hoofdeffect werd geclassificeerd door een interactie-effect: Zowel respondenten met een niet-duurzaam als met een duurzaam zelfbeeld rapporteren duurzame intenties wanneer zij zich nog niet duurzaam gedragen hadden (controle conditie). Echter, wanneer de respondenten zich duurzaam gedragen hadden (duurzame conditie), rapporteerden respondenten met een nietduurzaam zelfbeeld minder duurzame intenties dan respondenten met een duurzaam zelfbeeld. Deze studie laat zien dat vrijwaringseffecten vooral optreden wanneer mensen zichzelf als een nietduurzaam persoon zien en dat vrijwaringseffecten minder waarschijnlijk zijn wanneer mensen zichzelf als een duurzaam persoon zien. Ons onderzoek heeft belangrijke implicaties voor het stimuleren van duurzaam gedrag: voor herhaaldelijk duurzaam gedrag is het belangrijk dat mensen zichzelf als een duurzaam persoon zien. Eerder onderzoek toonde aan dat wanneer mensen na een duurzame gedraging expliciet gewezen werden op het duurzame karakter van deze actie, ze zichzelf als een duurzaam persoon gingen zien (Bem, 1967; Cornelissen et al., 2007). Ons onderzoek laat zien dat een dergelijke aanbeveling kan leiden tot een afname van vrijwaringseffecten, en daardoor tot meer effectieve (overheids) communicatie op dit terrein.
203
204
ETMAAL 2013
BOOK OF ABSTRACTS
Session #6 Thursday 7 February 16:45 – 17:45 hrs. Chair: Camiel Beukeboom Location: Auckland (M3-04) English/Dutch session
THE IRONY BIAS: HOW VERBAL IRONY REFLECTS AND MAINTAINS STEREOTYPIC EXPECTANCIES CHRISTIAN BURGERS AND CAMIEL BEUKEBOOM Imagine a teacher presenting an advanced mathematical problem to her class. A blond girl named Samantha convincingly provides the perfect answer. An impressed fellow student ironically yells “Wow, she is really dumb!” and the class laughs. Such an ironic comment may be seen as an arbitrary act of praise, that could just as well be formulated literally, by saying something like “Wow, she is really clever!”. However, the speaker’s irony use might also follow from existing stereotypic expectancies about the intelligence of blondes (Manning, 2010) that are violated by Samantha’s performance. Based on linguistic bias research, the question could be asked whether ironic comments result from, and maintain, stereotypic expectancies. Interpersonal communication plays is important in creating and maintaining social stereotypes. An extensive literature shows that speakers systematically vary their language when communicating stereotype-consistent and stereotype-inconsistent information (Beukeboom, in press). For instance, when communicating stereotype-inconsistent information, speakers not only use more concrete words (Maass et al., 1989; Wigboldus, et al. 2000), but also more explanations (Sekaquaptewa et al., 2003) and negations (Beukeboom et al., 2010) than when communicating stereotype-consistent information. These systematic variations not only reflect speakers’ stereotypic expectancies, but also strengthen them in both speakers and recipients. In this paper, we expand the literature on stereotype communication by introducing the Irony Bias. We posit that irony is an important linguistic means by which stereotypic expectancies are implicitly revealed and maintained. Linguists argue that irony is typically used in situations of violated expectancies (Utsumi, 2000; Wilson & Sperber, 2004). Ironic comments like Wow, she is really dumb always communicate two meanings: the literal (which in this example is negative) and the intended meaning (which in this example is positive). Both meanings remain active in working memory during and after processing (Akimoto et al., 2012; Giora et al., 1998). This makes irony particularly appropriate in case of violated stereotypic expectancies: irony allows speakers to literally mention the expectancy and, by the implied ironic meaning, indicate that this expectancy is violated in the current situation. Three experiments support our claim that irony is more appropriate in stereotype-inconsistent than stereotype-consistent situations. In experiment 1 (N=50, Mage=24.22, SDage=7.82, 54% female), participants were asked to choose between making literal and ironic comments in specific situations that varied according to a 2 (stereotype-consistent vs. stereotype-inconsistent) x 2 (positive vs. negative behavior) x 4 (stimulus set) within-subjects design. Results demonstrate that participants chose more irony in stereotype-inconsistent than in stereotype-consistent situations. 204
205
ETMAAL 2013
BOOK OF ABSTRACTS
Experiment 2 (N=58, Mage=30.48, SDage=13.25, 67% female) investigated the extent to which participants favor ironic and literal comments in stereotypic situations. Participants were asked to rate the appropriateness of literal and ironic comments in situations that varied according to a 2 (stereotype-consistent vs. stereotype-inconsistent) x 2 (positive vs. negative behavior) x 2 (literal vs. ironic comment) x 4 (stimulus set) within-subjects experimental design. Results demonstrate that literal comments were more appropriate in stereotype-consistent situations and that irony was more appropriate in stereotype-inconsistent situations. Findings were replicated in Experiment 3 (N=104, Mage=26.00, SDage=11.99, 62% female) in which behavioral valence (positive, negative) and stereotype-consistency (consistent, inconsistent) were between-subjects variables. Mediation analyses demonstrated that perceived behavioral expectedness mediated irony usage (experiment 1) and irony appropriateness (experiment 2). Experiment 4 focused on the communicative consequences of the irony bias. In line with other linguistic biases (Beukeboom, in press), we argue that the Irony Bias functions to implicitly communicate stereotypic expectancies to message recipients, and thus to contribute to the transmission and maintenance of socially shared stereotypes. In general, biased language use facilitates the transmission of essentialist beliefs about social categories, as measured in different inferences that recipients draw from biased target-behavior descriptions. In experiment 4 (N=87, Mage=26.51, SDage=11.60, 70% female), participants were asked to indicate their beliefs about a target person (i.e., perceived essentialism: repetition likelihood, dispositionality, generalizability) after reading literal or ironic comments about the target’s behavior. Descriptions varied according to a 2 (positive vs. negative behavior, within-subjects) x 2 (literal vs. ironic comment, within-subjects) x 4 (stimulus set, between-subjects) mixed experimental design. Results confirmed that perceived essentialism is lower after reading ironic than literal comments. Without any information about the actor, recipients inferred from ironic (compared to literal) comments about an actor’s behavior that the person was less likely to repeat the behavior in the future and in other situations, and that it was less dispositional and generalizable. In sum, we extend the stereotyping literature by showing that speakers reveal their stereotypic expectancies in ironic comments, and that such comments function to interpersonally communicate and maintain these expectancies.
COMMUNICATING BRANDS WITH AUDITORY CUES: INFLUENCE BRAND CHOICE WITH BRANDRELATED SOUNDS - GABY SCHELLEKENS AND PEETER W.J. VERLEGH Television and radio advertisements often use sounds to communicate brand cues. Consider for instance the ‘Ding-Dong’ of Domino’s Pizza Delivery or the popping of the lid of a Grolsch bottle. An association is created between a brand or spokes character of the brand (e.g., Energizer bunny, Tony the Tiger) and a specific sound effect (e.g., drumming, roar). Frequent exposure to the sound in combination with the brand are meant to create a learned association (i.e., evaluative conditioning; De Houwer, 2012). Exposure to the sound is supposed to automatically activate the brand. However, it remains unclear if a brand can be activated by an unconditioned auditory cue. So, when a sound that is semantically related to a brand spokes character, but is not previously played in 205
206
ETMAAL 2013
BOOK OF ABSTRACTS
advertisements, will exposure to this sound prime the brand of the spokes character? For instance, the sound of a cowbell is related to a cow in the Alps. Could exposure to the sound of a cowbell prime the brand Milka, the brand related to Alpine cows? The current study examined whether indirect reminders of brands by auditory cues associated with the spokes characters of brands affect consumer brand preferences. Previous research on auditory cues in marketing communication has mainly focused on music and jingles in television and radio advertisements. Research has shown effects of music on affect and moods (Alpert & Alpert, 1990; Bruner II, 1990;), waiting experience (Oakes, 2000), and the evaluations of a product, store or service (Garlin & Owin, 2006). The congruence (‘fit’) between music and an advertisement or brand has also been of main interest (Chebat, Chebat, & Vaillant, 2001). Compared to the extensive knowledge on music and jingles in marketing communication (Bruner II, 1990; Oakes, 2007), researchers know little about the effect of short sound effects in marketing communication. The study examines whether subtle, indirect reminders of brands by simple auditory cues (“sounds”) associated with the spokes characters of brands affect consumer preferences. The rationale for this idea stems from the classical work on spreading activation (Collins & Loftus, 1975) and priming (Higgins, Rholes, & Jones, 1977), which entails that the activation of mental representations activates semantically associated concepts. This activation can then exert an (unintended) and unconscious influence in subsequent contexts. Indirect brand triggers via auditory cues are also interesting from a marketing perspective, because they evoke a limited amount of consumer resistance (cf., Campbell, 1995; Kaikati & Kaikati, 2004). One hundred and eighty-three business students (age 19-25; 60% male) participated in this study. Participants had been randomly assigned to one of three conditions (cow bell vs. trumpeting elephant vs. no sound). The focal brands in this study are two brands of chocolate, Milka and Côte d’Or. Both brands are semantically associated with animal brand spokes characters. Milka uses a purple cow in much of its imagery and ads, and Cote d’Or uses a picture of an African elephant. In a filler task the sound of a cow bell, or a trumpeting elephant, or without any sound announced the end of a task. After this task participants were given the choice between a piece of Milka or Côte d’Or chocolate. Their choice was registered unobtrusively. From a funneled debriefing it became clear that none of the participants indicated any awareness of the goals or hypotheses of the study. This result made it less likely that the sounds led to a conscious thought process which affected the choice of chocolate. In line with our expectations the results of this study demonstrated that semantically related auditory cues can positively affect brand choice. A chi-square test revealed a significant effect of sound (bell vs. trumpeting elephant vs. no sound) on choice of chocolate (Milka vs. Côte d’Or): χ²(2) = 11.5, p = .003), see Figure 1. More specifically, exposure to the sound of a cowbell led to a preference for Milka chocolate (46 out of 70) compared to Côte d’Or, while hearing a trumpeting elephant led to preferring Côte d’Or (34 out of 61) over Milka chocolate. Pairwise comparisons show that the proportion of participants choosing Milka is higher after hearing a cow bell than after exposure to a trumpeting elephant (χ²(1) = 10.3, p = .001), or the (no sound) control condition (χ²(1) = 6.6, p = .01). Conversely, the preference for Cote d’ Or is higher after a trumpeting elephant than with a cow bell 206
207
ETMAAL 2013
BOOK OF ABSTRACTS
(χ²(1) = 11.1, p < .001), or the control condition (χ²(1) = 4.2, p = .04). The twenty participants who declined a piece of chocolate did not vary across conditions.
"BACKGROUND NOISE". A STUDY ON THE PROCESSING OF RADIO ADVERTISING WHILE MEDIA MULTITASKING - WENDALIN VAN DE GIESSEN, EDITH SMIT AND HILDE VOORVELD Historically, radio has always been seen as the less important and least rewarding medium in the media mix. But even in this new media age, radio is the second most used medium and consumed for 115 minutes radio per day (SPOT, 2010). Of those 115 minutes approximately two third is not spend solely on radio listening but is combined with other tasks (media multitasking). This is in line with other research which assumes that radio is consumed as a mainly secondary task (Perry-Husbands & Bowman, 2003). In this study it is tested how the “secondary nature” of radio influences the way radio commercials are perceived by consumers, and which implications this has for the persuasiveness of radio advertising. The following research question was formulated: “To what extent does the broadcasting of radio commercials while the receiver is attending another task have an effect on the recognition and the evaluation of those radio commercials?” We also took into account the congruence (‘fit’) between the secondary task performed and the commercial broadcasted and we formulated the following research question: “Do radio commercials with high congruence (‘fit’) with the secondary task enjoy a higher recognition and a more positive evaluation than radio commercials with low congruence (‘fit) between the commercial and the secondary task?” Based on the Elaboration Likelihood Model (Patty & Cacioppo, 1986), the Limited Capacity Model (Lang, 2000) and the Low Attention Processing theory (Heath, 2000), we hypothesized media multitasking would have a negative effect on recognition and recall, and a positive effect on evaluation and attitude. Furthermore, we expected that high congruence between the commercial and the task would result in higher recognition and evaluation of the radio commercial, based on the ‘automatic selection processing’ mechanism (Heath, 2000). To provide an answer to our research questions, we conducted two experiments. Both experimental set-ups consisted of a true 2 (frequency of broadcasting: zero vs. one, between-subjects) x 2 (task distraction: no task vs. task, between-subjects) x 2 (congruence/fit: high fit vs. low fit, withinsubjects) Factorial Design. The dependent variables in the first experiment were 1. Recognition of the radio commercials and 2. Evaluation of the radio commercials. The dependent variables in de second experiments were 1. Recall of the radio commercials 2. Recognition of the radio commercials 3. Evaluation of the radio commercials and 4. Attitude regarding the radio commercials. The first experiment stressed the importance of a strong external validity, whereas the second experiment stressed the importance of strong internal validity. The first experiment consisted of a total of 158 radio listeners with a mean age of 49 years, and was conducted in cooperation with the Radio Advies Bureau (RAB). The second experiment was conducted in a lab-setting at the Comlab of ASCoR. The sample of experiment (n=118) consisted of mainly students (98,3%), with a mean age of 21 years. The results of the first experiment showed a significant lower recognition of radio commercials when the participants were attending to another task (thus media multitasking). Moreover, the results also 207
208
ETMAAL 2013
BOOK OF ABSTRACTS
showed a significant higher evaluation of the radio commercials when the participants were attending to another task. The commercials which enjoyed a high congruence with the task were significantly more recognized and significantly higher evaluated, than the commercials with a low congruence. The results of the second experiment were not completely in line with the results of the first experiment. The results of the second experiments showed a significant higher recognition of radio commercials while the participants were attending to another task, which is the opposite of the result found in the first experiment. Furthermore, unlike the results of the first experiment, there were was no main effect found on evaluation and commercials with high congruence were not recognized significantly more than commercials with low congruence. The result that was in line with the first experiment was that the commercials with high congruence were significantly more positive evaluated than the commercials with low congruence. A possible explanation for the differences found is the difference in sample; younger people might have a higher media multitasking capacity than older people (Carrier et al. 2009). We can conclude that 'fit' is an important aspect in creating a positive attitude regarding an radio advertisement, and that radio is a powerful medium in creating advertising effects, regardless of its “secondary nature”.
POSITIEVE EN NEGATIEVE UITINGEN: EEN KRACHTSTRIJD - CHRISTINE LIEBRECHT, LETTICA HUSTINX AND MARGOT VAN MULKEN Een schrijver kan zich ergens positief of negatief over uitlaten en hierin mild of zeer uitgesproken zijn. Wil hij zijn evaluatieve uiting krachtig verwoorden, dan kan hij kiezen uit een groot palet formuleringsalternatieven. Een van de mogelijkheden is het gebruik van taalintensivering, een stilistisch middel waarmee oordelen kracht bijgezet worden. Een uiting kan bijvoorbeeld variëren van gematigd en licht positief tot sterk positief. Vergelijk 'het hotel was goed' met 'het hotel was geweldig'. Hetzelfde geldt voor negatieve uitingen ('slecht' versus 'verschrikkelijk'). In de loop der jaren zijn er diverse corpus- en effectonderzoeken uitgevoerd naar taalintensivering (zie voor een overzicht Van Mulken & Schellens, 2012). Echter, de polariteit (ook wel: valentie) van een uiting is hierbij niet betrokken. Dit is opvallend, aangezien positieve en negatieve boodschappen een andere werking hebben op ontvangers: negatieve boodschappen en woorden komen krachtiger over dan positieve, toonden eerdere onderzoeken aan (o.a. Baumeister e.a., 2001; Feldman, 1966). Een tweetal theorieën verklaart deze krachtigere werking van negatieve boodschappen. Ten eerste zijn negatieve boodschappen krachtig op cognitief niveau: mensen hebben van nature automatisch de neiging om aanzienlijk meer aandacht te besteden aan onaangename (negatieve) dan aan aangename (positieve) informatie (de zogeheten negativity bias) (Jing-Schmidt, 2007; Parkinson, 1995; Pratto & John, 1991). Ten tweede zijn negatieve boodschappen krachtig op sociaal niveau. Taalgebruikers hebben de universele neiging om zaken vanuit de zonnige kant van het leven te bekijken en erover te vertellen (dit wordt het Pollyanna principe genoemd) (Boucher & Osgood, 1969; Jing-Schmidt, 2007). Is een boodschap negatief, dan wordt de verwachting geschonden en komt deze dus krachtiger over.
208
209
ETMAAL 2013
BOOK OF ABSTRACTS
Op basis van dit verschil in verwerking van positieve en negatieve boodschappen, is te verklaren dat een negatief geformuleerde boodschap een grotere kracht heeft dan een positief gestelde boodschap. Met andere woorden: een half leeg glas is, in de perceptie van de taalgebruiker, leger dan een half vol glas. De vraag is welke rol taalintensivering speelt bij de gepercipieerde kracht van positieve en negatieve evaluaties. Van dit stilistische middel wordt gesteld dat het uitingen krachtiger maakt, maar nog niet eerder is systematisch het verband onderzocht tussen taalintensivering en de polariteit van de uitingen. Is een 'heel erg half leeg glas' leger dan een 'heel erg half vol glas' vol is? Er is een experiment opgezet om de vraag te beantwoorden in hoeverre taalgebruikers een krachtverschil ervaren tussen al dan niet geïntensiveerde positieve en negatieve uitingen. Evaluaties kunnen betrekking hebben op allerlei dimensies, zoals schoonheid of interessantheid. Voor dit onderzoek zijn 5 dimensies onderscheiden: schoonheid (is iets mooi of lelijk), intelligentie (is iemand slim of dom), smaak (is iets lekker of vies), interessantheid (is iets interessant of saai) en kwaliteit (is de kwaliteit van iets goed of slecht). In een pretest kregen participanten per dimensie alleen het positieve of negatieve woord te zien. Daarbij dienden zij een antoniem en twee intensiveerders van het gegeven woord op te schrijven. De woorden die het meest frequent door de participanten waren genoemd als antoniem en intensivering zijn meegenomen in het materiaal van het experiment. Zo ontstond er bij iedere dimensie een kwartet van woorden: een ongemarkeerd positief en negatief woord (bijvoorbeeld 'mooi' en 'lelijk') en een geïntensiveerd positief en negatief woord ('prachtig' en 'afschuwelijk'). In totaal bestond het materiaal dus uit 20 items (5 (dimensies) x 4 (condities)). Deze werden in een binnenproefpersoonontwerp aan de participanten voorgelegd. Eerste analyses laten zien dat negatieve uitingen inderdaad krachtiger overkomen op taalgebruikers dan positieve uitingen. Dit was zowel bij de ongemarkeerde uitingen het geval ('lelijk' > 'mooi') als bij de geïntensiveerde uitingen ('afschuwelijk' > 'prachtig'). Aan de negatieve zijde was de afstand tussen een ongemarkeerde uiting en een geïntensiveerde uiting kleiner dan aan de positieve zijde. Deze resultaten wijzen erop dat positiviteit inderdaad default is, zoals het Pollyanna principe pretendeert, en dat negativiteit inderdaad krachtiger binnenkomt bij ontvangers, zelfs als de uiting is geïntensiveerd. Een 'half leeg glas' en een 'heel erg half leeg glas' zijn dus leger dan een '(heel erg) half vol glas'.
Session #7 Friday 8 February 09:00 – 10:00 hrs. Chair: Guda van Noort Location: Rochester (M2-10) English session
DISCLOSING ONLINE BEHAVIORAL ADVERTISING: EFFECTS OF THE COOKIE-ICON - GUDA VAN NOORT, EDITH SMIT AND HILDE VOORVELD New Internet technologies provide the possibility of automated tracking of consumers’ Internet behaviour. Such tracking is used to create user profiles for the purpose of displaying advertisements that fit the interest of these individuals (e.g., Goldfarb & Tucker 2011). Confident in the success of 209
210
ETMAAL 2013
BOOK OF ABSTRACTS
such targeted ads, businesses invested heavily in online targeted advertising and this spending is expected to increase. However, this marketing strategy, also referred to as “online behavioural advertising” (OBA) is heavily debated with regard to the protection of user privacy. In response to (upcoming) governmental privacy regulations to protect consumers, the Dutch industry introduced the “cookie-icon” as a form of self-regulation. This cookie-icon is displayed in online behavioural advertisements and aims at warning Internet users that particular advertisements are based on their past surfing behaviour. The first aim of this study is to investigate whether consumers are aware of this cookie-icon and to investigate their understanding of the meaning of this icon. Theoretically, the cookie-icon can be interpreted as disclosure or forewarning. Based on forewarning literature it is expected that the cookie-icon is only effective when the user has existing knowledge of the icon and OBA as a marketing technique (i.e., persuasion knowledge). Only then, users might understand what the cookie-icon informs them about, which enables them to cope with the information and respond to it in an effective manner (e.g., Friestad & Wright 1994). Examples of such coping strategies would be adjusting cookie preferences or avoiding behavioural targeted advertisements. The second aim of our study is to examine whether the cookie-icon is able to influence such coping strategies. Since the cookie-icon is relatively new, we suggest that individuals do not have sufficient levels of knowledge for the cookie-icon to be effective as a warning and claim that a short explanatory label (i.e., “this ad is based on your surfing behaviour”) should accompany the icon to be effective. Moreover, research demonstrated that forewarnings help consumers to cope with persuasive messages. As Wood and Quinn (2003) stated “forewarned is forearmed”, meaning that if users know the persuasive intent of a message, they will be more difficult to persuade. Therefore, the third aim of the study is to examine whether displaying a forewarning such as a cookie-icon, with or without explanatory label, indeed results in less persuasion. Prior studies indicate that consumers lower their brand and ad evaluations after learning about controversial marketing tactics companies use to persuade them (Lee 2010; Milne, Rohm & Bahl 2009; Wei, Fisher & Main 2008). Therefore, it could be expected that disclosing targeted ads by means of the cookie-icon negatively influences attitudes toward the brand and the online ad. Negative evaluations of the advertisement might also transfer to the evaluation of website on which the disclosure is presented. An experiment (N = 326, Mage = 33.71, 58% female) with three experimental conditions (no cookieicon, cookie-icon only, and cookie-icon with short explanatory label) tested the effectiveness of the cookie-icon as a disclosure of online behavioural targeted advertising in terms of cookie awareness and understanding, coping strategies and persuasion effects (i.e., brand, ad and website responses). The results showed several intriguing findings. It was demonstrated that the cookie-icon, whether or not accompanied by an explanatory label, was noticed by consumers. However, the cookie-icon in itself was insufficient in activating awareness of behavioural advertising. The explanatory label was needed to activate this knowledge. In turn, this knowledge engendered coping strategies to deal with behavioural advertising. However, brand, ad and website responses were not negatively influenced by the cookie-icon. Actually, brand recall and perceived relevance of the brand and ad increased when the cookie-icon was presented.
210
211
ETMAAL 2013
BOOK OF ABSTRACTS
The results of this study implications for the self-regulation initiative of the industry. Most important, the current study showed that adding an explanatory label increased the understanding of the cookie-icon, while this did not negatively spill over to the evaluations of the brand, ad or website. Thus, it can be questioned whether the cookie-icon is a good way to forewarn consumers about OBA since it seems that consumers do not use their persuasion knowledge to protect them from being persuaded. The study has also some interesting implications for advertisers and website providers. For advertisers, results of the current experiment were very positive. Adding a cookie-icon (with or without explanatory label) to their targeted ads positively influenced brand recall, perceived ad relevance and brand relevance. Finally, website providers should not resist cooperating with this selfregulating initiative as the cookie-icon did not influence attitude towards the website and perceived relevance of the website.
PROTECT YOURSELF… IF YOU CAN: IS PROTECTION AGAINST SUBLIMINAL ADVERTISING POSSIBLE? - STEFAN F. BERNRITTER, PAUL KETELAAR, THIJS VERWIJMEREN, JOHAN C. KARREMANS, WOLFGANG STROEBE AND DANIËL H. J. WIGBOLDUS An emerging body of research indicates that subliminal advertising is effective if the advertised product is goal relevant. These findings revive the discussion whether we are at the mercy of this unconscious persuasion technique. Therefore, in the present two studies, we investigated whether there are possibilities for people to protect themselves against subliminal advertisements. It is assumed that the effects of subliminal advertising rely on the activation of automatic responses. Consequently, we examined whether making people conscious about the fact that they are exposed to subliminal advertisement by means of a warning message could neutralize those automatic responses. Furthermore, we investigated whether the moment at which a warning is presented affected the effectiveness of subliminal ads. Findings revealed that it is possible to protect people from subliminal ads by showing them a warning message before being exposed to subliminal ads. Participants who were warned after exposure to subliminal ads (i.e., just before they had to make a choice between two beverage brands, of which one has been subliminally primed before), and participants who were not warned at all did not show this effect. Moreover, different strategies to protect people from subliminal advertisements and implications for future research are discussed.
ONLINE BEHAVIOURAL ADVERTISING: DO WE KNOW HOW TO COPE? - EDITH SMIT, HILDE VOORVELD AND GUDA VAN NOORT Online behavioural advertising (OBA) is a form of targeted advertising for which data about online surfing behaviour are being collected, usually by installing “cookies”. Cookies are small text files that are put on users’ devices in order to facilitate the functionality of that specific website (i.e., first party cookies) or to collect profile information for targeted advertising (i.e., third party cookies). The use of cookies is heavily debated within the European Union, because of its potential violation of the privacy of Internet users. In the new EU e-Privacy directive it is stated that data storage by placing cookies is only allowed if the user has been well informed and gave permission (i.e., informed consent). The previous “opt-out” possibility is no longer sufficient. Websites nowadays should explicitly ask permission for using personal data through so called “opt-in”. This new legislation requires that website users have active control and executing this control requires knowledge about 211
212
ETMAAL 2013
BOOK OF ABSTRACTS
OBA. With our study we aim to answer the question what Dutch Internet users know about these practices and how they cope with OBA and cookies. An online questionnaire was fielded briefly after the new Dutch cookie regulations were effectuated in June 2012 (N = 2022, 49.50% female; Mage = 45.28). Results showed that respondents had a pretty good understanding of OBA, with on average five out of eight statements (McDonald & Cranor, 2010) being correctly answered. Respondents most often were mistaken about the statement that organizations are not allowed to gather and store information about individual’s Internet behaviour. 58.6% thought it was true, while it is in fact not true. Knowledge about cookies was also measured with eight true/untrue statements (McDonald & Cranor, 2010). On average they had four out of eight statements correct, which is significantly lower than the OBA knowledge. Most misunderstanding was about the idea that cookies save your history, that cookies are person-based, and that a computer will slow down when cookies are not regularly removed. Respondents were concerned about their online privacy (Baek & Morimoto, 2012) with an average score of one point above the midpoint of the 7-point agreement scale. They were especially worried about the misuse of their personal data. We distinguished different ways of coping with OBA, namely resistance strategies (Jacks & Cameron, 2003) toward OBA, online privacy protection, and willingness to accept cookies (i.e., regarding first and third party cookies). Overall respondents were not very likely to resist (M = 4.2 on 7-point scale). Inspection of the items indicate that social validation is the least likely coping strategy, and that selective exposure and asserting confidence are the most likely coping strategies. There was a general intention to protect one’s privacy (M = 3.62 on a 5-point scale), most often by checking spy ware, block pop-ups and removing their browser history. Respondents less often remove cookies. Reading privacy statements is the activity they do least often which is remarkable because this is the activity that is asked for in the privacy regulations. After explaining what cookies are nearly 6% was willing to accept all type of cookies, 35% was willing to accept first party cookies , only 7% was willing to accept third party cookies, 37% was not willing to accept any cookies, and 16% didn’t know if they would accept cookies. When comparing website types, respondents were least willing to accept cookies for corporate websites of profit or non-profit companies (45% and 43%, respectively) whereas for other websites (i.e., news websites, social network sites, email services, search engines, and governmental websites) around 35% was not willing to accept any cookies. In conclusion, the level of cookie knowledge is not very high amongst our sample of heavy Internet users. We expect that the knowledge in the general population is even lower. Our sample was concerned about their privacy. They were especially worried about misuse of their personal data. Some coping strategies were more often used than other strategies. Overall, respondents showed some resistance strategies and some adjustments of their settings. Regression results showed that those who know about cookies and those who are less concerned, are the ones that are more active in online privacy protection and more willing to accept first party cookies. Our survey raises a fundamental question in the online privacy protection discussion, namely: what does informed consent mean within a not-informed audience? Internet users should first be better informed before legislation can be applied. We suggest a dual approach here: not only inform users (more and better), but also to reduce concerns about online privacy.
212
213
ETMAAL 2013
BOOK OF ABSTRACTS
DIFFERENCES BETWEEN YOUNGER AND OLDER ADULTS IN RESPONSES TO EMOTIONAL TELEVISION COMMERCIALS - MARGOT VAN DER GOOT, EVA VAN REIJMERSDAL AND MARISKA KLEEMANS Introduction People aged 50 years and older are widely recognized as an important target group for commercial organizations (Yoon, Cole & Lee, 2009), which renders insight into the specifics of this age group much-needed. Based on psychological research, we can expect that older consumers are different from younger consumers in the way they respond to advertising. Research, supporting the socioemotional selectivity theory (e.g. Carstensen, Isaacowitz, & Charles, 1999), shows that older adults differ from younger adults in their emotional goal setting in life, suggesting that these groups also differ in their responses to emotional advertising. However, academic advertising research focusing on age differences is scarce. Therefore the current study aims to investigate whether there are age differences in liking and recall of emotional versus rational television commercials. Theoretical background Socioemotional selectivity theory argues that older adults have a heightened sense of time being limited, which makes them seek out pleasurable, “gentle”, experiences, whereas younger adults perceive time as more open-ended and therefore focus on optimizing the future by acquiring new information (Carstensen, Fung, & Charles, 2003). Two studies, which derived hypotheses from this theory, found that older people liked print advertisements with an emotionally meaningful appeal better than advertisements with a rational appeal (Fung & Carstensen, 2003; Williams & Drolet, 2005), whereas younger adults did not show this bias (Fung et al., 2003) or showed higher liking for rational ads (Williams & Drolet, 2005). Also, older adults remembered emotional content better than rational content because it was more relevant for them (Fung & Carstensen, 2003). We expect that the same mechanism applies to television commercials. Therefore, we hypothesize that older adults like commercials with emotionally meaningful content more than rational commercials and remember brands in emotional commercials better, whereas this difference would not occur for younger adults. Method An experiment was conducted, using a 2 (emotional versus rational commercial) x 2 (old versus young adults) mixed design. The stimulus material consisted of six commercials: three emotional commercials and three rational commercials. Emotional commercials showed love, warmth, and families, whereas rational commercials were selected based on their informative and factual content (Fung & Carstensen, 2003). Commercials were chosen in three product categories that were equally relevant for both age groups. A manipulation check showed that self-reported emotional arousal was indeed significantly higher for the emotional commercials than for the rational commercials. A total number of 83 participants took part in the experiment. The older age group (N= 39) consisted of people aged 50 years and older (M = 58.28; SD = 7.75) and the younger group (N = 44) consisted of people with ages between 20 and 30 years (M = 25.09; SD = 2.57). Participants watched all six 213
214
ETMAAL 2013
BOOK OF ABSTRACTS
commercials, and indicated after each commercial their liking of the commercial by means of four 7point items that indicated how favorable, enjoyable, likeable, and good they perceived the commercial. After watching all commercials, participants were asked which brands they had seen in the commercials (unaided recall). A repeated measures MANOVA was carried out, with emotional content as within-subject factor and age as between-subjects factor. Results and discussion As expected, we found a significant interaction between emotional content and age for ad liking. Older adults liked the emotional commercials (M = 5.08; SD = 0.15) significantly more than the rational ones (M = 3.90; SD = 0.11). The same applied to the younger adults (emotional content: M = 4.55; SD= 0.14; rational content: M = 4.15; SD = 0.11), but for them the difference between the two types of commercials was smaller. Our hypothesis for recall was not supported: We did not find an interaction between emotional content and age. We did find main effects of emotional content and of age: recall was higher for emotional than for rational commercials, and older adults remembered the brands less well than the younger adults. In sum, this study is the first to show that older adults like emotional television commercials better than rational ones, and that older adults like emotional commercials better than younger adults. These findings speak to the ability of the socioemotional selectivity theory to predict age differences in responses to television advertising. Practitioners are recommended to include warm and pleasant content when targeting older adults. The reason that we did not find the hypothesized effect for recall could be that older adults remember less from commercials than younger adults because of the audiovisual production features of television commercials regardless of the type of content, thus ‘overriding’ the positive effect of emotional content.
Session #8 Friday 8 February 10:15 – 11:15 hrs. Chair: Eva van Reijmersdal Location: Forumzaal (M3-15) English session
CHILDREN’S RESPONSES TO ADVERTISING IN SOCIAL GAMES: PERSUASION KNOWLEDGE AND SUSCEPTIBILITY TO PEER INFLUENCE - ESTHER ROZENDAAL, EVA VAN REIJMERSDAL, NOORTJE SLOT AND MONIEK BUIJZEN In countries all over the world, children spend increasingly more time playing social games. Social games are online communities where children can meet and interact with new and existing friends and play games. Some of the most popular social game communities, such as Neopets, Club Penguin and Habbo attract millions of young visitors each month.These games are popular not only among children but also with advertisers. Social games provide a unique environment for advertisers to interact with and engage young consumers. However, despite its fast-growing popularity, little is known about how children perceive and respond to advertising in social games. This study aims to take a first step in gaining a better understanding of this increasingly popular phenomenon. 214
215
ETMAAL 2013
BOOK OF ABSTRACTS
We focus on two defining characteristics of advertising in social games. The first characteristic is the high level of integration between advertising and the game context. Advertising in social games is integrated within the game, for example, brands appear on billboards and on virtual goods or even have their own ‘branded’ environment as part of the online community. It is generally assumed that children have difficulty recognizing and understanding the commercial nature of these integrated advertising practices, which may result in stronger persuasion. Therefore, we are specifically interested in the level of children’s knowledge concerning the advertising integrated into social games (i.e., persuasion knowledge) and how that knowledge is related to their advertising susceptibility. The second defining characteristic of advertising in social games is its social nature. In social games, advertisements are placed in a context in which peers play a central role. Peers (e.g., friends and classmates) have long been recognized as playing a socialization role in shaping young people’s consumer behavior. As children develop into teens, they start to believe that the possession of certain products and brands determines the quality of their friendships, and therefore, their willingness to conform to the peer group is higher. In this way, the peer group sets the standard for which products are cool and popular. It is likely that peer influence also occurs in social game environments, which may make children more susceptible to the persuasive effects of the advertising that they encounter in these games. Thus, the aim of this study is to explore children’s level of persuasion knowledge and peer influence susceptibility concerning advertising in social games and to investigate how these variables affect children’s desire for the brands advertised in these games. We focus on 9- to 12-year olds, because the children in this age group are most interested in playing social games and their willingness to conform to the peer group is at its peak. The sample consisted of 148 children of 9 to 12 years old (57% boys). The study focused on Habbo, one of the world’s largest social games for children and teens. Because of the study’s exploratory nature, a descriptive correlational design was used. Exposure to advertising in Habbo was controlled and equal for all children. Persuasion knowledge (i.e., advertising recognition, understanding commercial source, understanding persuasive intent, critical advertising attitude) and peer influence susceptibility were measured as independent variables, and advertised brand desire as the dependent variable. Results showed that (1) children have a fairly good knowledge of the advertising in social games, yet hold uncritical attitudes towards it and claim to be not very susceptible to peer influence, (2) the most important predictors of children’s desire for the advertised brands are a low critical attitude and high peer influence susceptibility, and (3) recognition and understanding of advertising in social games is only effective in reducing advertised brand desire for children who are familiar with social games. These results have important implications for the on-going societal and political debate about children and advertising, which will be discussed.
215
216
ETMAAL 2013
BOOK OF ABSTRACTS
EFFECTS OF PROFILE TARGETING ON CHILDREN’S RESPONSES TO ONLINE ADVERTISING - EVA VAN REIJMERSDAL, ESTHER ROZENDAAL, NADIA SMINK, GUDA VAN NOORT AND MONIEK BUIJZEN Both adults and children post information on their own profile pages on social network sites about their preferences for brands, food, music, hobbies, etcetera. By means of so-called profile targeting, advertisers use this information to tailor their campaigns to the interests of individual consumers (Villiard & Moreno, 2012). This practice has raised concerns in society because personal information is used by the advertiser for persuasive purposes (Barnes, 2006; Mahr, De Ruyter, Wetzels, & Grewal, 2012). In this study, we focus on the effects of profile targeting on children’s cognitive, affective, and behavioral responses to online advertising and we examine three factors that may explain these effects, 1) perceived ad relevance, 2) perceived targeting, and 3) positive attitudes toward the targeted advertising. First with respect to ad relevance, previous research among adults has shown that online targeted advertising can increase the persuasiveness of campaigns because they are perceived as personally relevant (Antheunis & Van Noort, 2012; Tam & Ho, 2006). Limited capacity models of information processing (Buijzen, Van Reijmersdal, & Owen, 2010; Lang, 2000), predict that if a message is relevant, people are motivated to allocate more cognitive resources which increases attention to the message. It is expected that targeted ads are perceived as more personally relevant, because they are based on people’s interests, which leads to more persuasion. Second, with respect to perceived targeting, research has also shown negative effects of online targeted advertising. When the ad becomes too personal, adults realize that their information is used to target the advertisement, resulting in perceptions of intrusiveness and vulnerability which negatively affect responses to the ad and the brand (Hoekstra & Van Doorn, 2012; Mahr et al., 2012). However, children lack the knowledge and skills to process online targeted advertising critically. They may not perceive the ad as targeted or intrusive, which makes them more vulnerable to persuasion (Rozendaal, Lapierre, Van Reijmersdal, & Buijzen, 2011). Third, ad liking may explain effects of targeted ads on children. Although targeted advertising has proved to gain attention online, research has shown that online banner ads in general receive very little attention and are processed on a heuristic or automatic level (Drèze & Hussherr, 2003; Yoo, 2008). On these levels of processing, affective mechanisms play an important role especially among children because their cognitive capacities have not fully matured (Buijzen et al., 2010). Therefore, we may expect that effects of online profile targeting on affective and behavioral responses to the brand are caused by ad liking. A total of 231 children of 8 to 13 years old (M = 10.82, SD = 0.80, 51.7% female) participated in the research. This study employed a 2 (color targeted) x 2 (product targeted) between-subjects design. Based on questions about children’s favorite color and leisure time activity, children were randomly assigned to a condition with a banner ad in their favorite color or not in their favorite color and/ or for a product that fitted their favorite leisure time activity or that did not fit their favorite activity. 216
217
ETMAAL 2013
BOOK OF ABSTRACTS
The banner ad was placed on a personal webpage of the Dutch social network site Hyves of a fictitious child with a gender neutral name. The results showed that targeting online advertising toward children’s favorite hobby leads to more positive brand attitudes and increased purchase intention. There were no effects on cognitive responses; Children did not recall the targeted ads more often than the ads that were not targeted. There were also no effect of color targeting. With respect to the underlying mechanisms, the analyses showed that children did not understand that the ads were targeted. Similarly, they did not find the targeted ads more relevant. The only process that did explain the positive effects of targeting was affective in nature; Children like the targeted ads more, which resulted in more positive attitudes and increased purchase intentions. This study is the first to show that online ads that are targeted based on SNS profile information lead to more positive brand attitudes and increased purchase intentions among children. These effects cannot be explained by increased perceived relevance or perceived targeting, but are the result of an affective mechanism that drives attitudes and intentions through ad liking. Unlike adults, children do not process targeted advertising in a cognitive and critical manner, rather they are influenced on a heuristic or automatic level via affect. This is also illustrated by the lack of effects on cognitive responses. These findings have important theoretical and practical implications.
CAPTURING CHILDREN’S ADVERTISING EXPOSURE: A COMPARISON OF METHODS AND MEASUREMENTS - SUZANNA J. OPREE, MONIEK BUIJZEN, EVA VAN REIJMERSDAL AND PATTI M. VALKENBURG Today’s Western children grow up in a heavily commercialized media environment. Advertising is everywhere: children are bombarded with advertising from the minute they wake up to the moment they go to sleep (Levin & Linn, 2004). The expenditures on children’s advertising have been increasing for decades, and so have the concerns about its potential side effects (Kunkel et al., 2004). Children are believed to be particularly vulnerable to the effects of advertising, and parents, psychologists, and policy makers worry that advertising contributes to children’s unhappiness, poor eating habits, and negative self-perceptions (Smith & Atkin, 2003). The method most appropriate for studying the effects of advertising depends on the outcome variable of interest. Experiments, for example, can be used to investigate the immediate effects of advertising exposure on product desires and product purchase requests. However, the more subtle, long-term effects of advertising are best to be investigated by means of longitudinal survey and/or media diary studies as changes in happiness, eating habits, and negative self-perceptions don’t occur overnight (Harmon, 2001). Yet, only few scholars have conducted longitudinal survey and/or media diary studies into the effects of advertising. Because existing studies use different measures for assessing children’s advertising exposure, their results are difficult to compare (cf. Buijzen & Valkenburg, 2003). To enhance knowledge accumulation, a standard measure of advertising exposure is called for. This study aims to compare existing measures in order to determine which one is best. In the current day and age, 8- to 18-year-olds spend 4.5 hours per day watching television and 1.5 217
218
ETMAAL 2013
BOOK OF ABSTRACTS
hour per day using the Internet (Rideout, Foehr, and Roberts, 2010). For that reason we will focus on measuring both children’s exposure to television advertising and exposure to Internet advertising. Studying the existing literature on children’s advertising exposure and media exposure, we have derived eight self-report measures for each. Four of these measures are survey measures and four are media diary measures. They measure respectively (1) exposure to the medium, (2) exposure to media titles, (3) exposure to commercial media titles, and (4) exposure to media titles, weighted for the advertising density of those media titles. ‘Exposure to the medium’ refers to either television or the Internet. The media titles we used are television shows (i.e., for measuring television advertising exposure) and websites (i.e., for measuring Internet advertising exposure) popular among the age group under study. Data on the advertising density of the television shows and websites were provided by Nielsen Market Research. In general, it is believed that estimates of time use that are obtained through media diaries are more accurate than those obtained by means of survey research. Also, the predictive value of measures is believed to increase with their specificity (Romantan et al., 2008; Slater, 2004). To examine the applicability of our advertising exposure measures, we wanted to determine their test-retest reliability and construct validity (DeVellis, 2003). To this end, we collected longitudinal data at schools among 166 8- to 11-year-olds. The children completed two identical surveys with a two-week interval. Of these children, 90 were also requested to keep media diaries. These diaries were to be completed on two weekdays and one weekend day subsequent to the administration of the questionnaire. Prior to the implementation of the study, informed consent from both children and parents were obtained. Test-retest reliability was assessed by looking whether children’s responses were consistent over time. All sixteen measures performed excellent to this respect. For the construct validity we examined whether children’s responses to the various measures were correlated to each other and whether their responses were correlated to other measures (gender, age, advertised product desire, parent-child conflict, and materialism) in expected ways. Contrary to expectations, the construct validity was better for the survey measures than for the media diary measures. In addition, the construct validity was better for ‘simple’ measures than for more specific measures. In our full paper and presentation we will elaborate on the theoretical explanations and implications of these results. For now, we conclude that children’s exposure to television advertising can best be measured by asking them (1) how often they watch certain television in the morning / afternoon / evening or (2) how often they watch popular shows on commercial networks. Similarly, children’s exposure to Internet advertising can be measured by asking children (1) how often they use the Internet in the morning / afternoon / evening or (2) how often they visit certain popular websites.
ENGAGING ENTERTAINMENT-EDUCATION: HOW ENGAGEMENT IN ENTERTAINMENT-EDUCATION NARRATIVES RELATES TO UNFAVORABLE THOUGHTS AND EFFECTS - LONNEKE VAN LEEUWEN, BAS VAN DEN PUTTE, REINT JAN RENES AND CEES LEEUWIS Entertainment-Education (E-E) narratives contain implicit educational messages to positively influence determinants of prosocial or health behavior (Kaiser Family Foundation, 2004). Following the E-E strategy, the Trimbos Institute co-developed a televised dramaseries called Roes (“High”) that aimed to reduce binge drinking (BD) among youngster. A previous study showed that Roes 218
219
ETMAAL 2013
BOOK OF ABSTRACTS
successfully influenced alcohol use determinants (Van Leeuwen, Renes, & Leeuwis, in press)). Though Roes and other E-E narratives have shown to be effective, the psychological mechanisms behind the effects of E-E narratives are not well understood (Moyer-Gusé, 2008; Dal Cin, 2004). Therefore, the present study explores the mechanism behind the effects of E-E narratives. A shortened version of a Roes episode is employed in this study. We expect (H1) that exposure to this episode is associated with desired effects on BD evaluation. Engagement in narratives is theorized as a psychological mechanism mediating the effects of E-E narratives (e.g. Green & Brock, 2000; Slater & Rouner, 2002). While engaged in E-E narratives, audience members emotionally connect to characters, feel present in the narrative world and lose their self-awareness. Our goal is to examine the hypothesis (H2) that narrative engagement is associated with unfavorable BD evaluations. Furthermore, unfavorable thoughts about the E-E narrative should divert attention from following the E-E narrative which interferes with narrative engagement (Busselle & Bilandzic, 2008). An aspect that may trigger unfavorable thoughts is the quality of the narrative. Therefore, we hypothesize (H3) that unfavorable thoughts about the quality of an E-E narrative are negatively associated with narrative engagement. Method The study employed an online pre (T0) and posttest (T1) between-subjects design. All participants (N = 176, Mage = 20, 84.1% female, 15.9% male) viewed a shortened version of an episode of Roes, portraying how a female protagonist experiences negative consequences of BD. Participants were randomly assigned to three conditions: standard version (n = 64), audiovisual disturbances version (n = 55) and critical instruction (n = 57). In the audiovisual disturbances version, quality was manipulated by adding beeps and stills, aiming to suggest that the episode had been badly edited. In the critical instruction condition, participants were instructed to critically evaluate the quality of the standard version. These two manipulations were intended to trigger unfavorable thoughts about the quality of Roes. The T0 questionnaire was completed at least one week before viewing the episode and included questions about demographics and BD evaluation (risk perceptions, perceived benefits, general attitude towards BD, perceived vulnerability to risks and perceived severity of risks). The T1 questionnaire was completed directly after viewing the episode and included questions on three dimensions of narrative engagement (narrative understanding, attentional focus, experiential engagement), a self-report on unfavorable thoughts about the narrative and BD evaluation items. Preliminary results H1: Paired samples T-tests on the BD evaluation items at T0 and T1 revealed that risk perceptions were higher (p < .001), general attitude was more negative (p < .001), and perceived vulnerability was higher (p < .01) at T1. In contrast to our expectations, perceived benefits of BD were also higher (p < .01). Perceived severity on T1 did not differ from perceived severity on T0. H2: Regression analyses were conducted on the BD items at T1 with the following predictors: BD measures at T0, gender, narrative understanding, attentional focus and experiential engagement. Results showed that attentional focus predicted perceived vulnerability negatively (β = -.15, p < .05) and perceived severity positively (β = .25, < .01) at T1. Experiential engagement positively predicted perceived vulnerability at T1 (β = .27, p < .001) and risk perceptions (β = .24, p < .01) at T1. The three narrative engagement dimensions did not predict general attitude at T1. 219
220
ETMAAL 2013
BOOK OF ABSTRACTS
H3: A one-way ANOVA showed that participants in the critical instruction and audiovisual disturbances conditions produced significantly more unfavorable thoughts about the narrative than those in the standard version condition. Regression analyses revealed that unfavorable thoughts negatively predicted narrative understanding (β = -.27, p < .001), attentional focus (β = -.31, p < .001) and experiential engagement (β = -.36, p < .001). Conclusions The results of this study confirm H1 that an E-E narrative is able to affect BD evaluation. The results also confirm H3 that audience members’ critical thoughts about E-E narratives decrease narrative engagement (i.e., attentional focus, narrative understanding and experiential engagement). Because, supporting H2, experiential engagement and attentional focus influence BD evaluation, critical thoughts about the quality of an E-E narrative may indirectly influence the effects of the E-E narrative.
Session #9 Friday 8 February 10:15 – 11:15 hrs. Chair: Marjolijn Antheunis Location: Rochester (M2-10) English/Dutch session
CREDIBILITY IN E-WOM: HOW REVIEW PERCEPTIONS IMPACT THEIR PERSUASIVENESS - NATALIE VAN HEMELEN, PEETER VERLEGH AND TIM SMITS For a long time, firms have understood that word-of-mouth is a very strong and influential advertising tool that can ‘make or break’ the image of a product or service. With the advent of social media, electronic word-of-mouth (e-WOM) and online consumer reviews (i.e. online recommendations about products, services, organizations or brands, based on a consumer’s personal experience) became increasingly popular (e.g., Willemsen, Neijens, Bronner & De Ridder, 2011). According to the Local Consumer Review Study (2012) most people trust online reviews and don’t doubt about their authenticity (Anderson, 12.05.2012). This suggests that reviews probably have a more persuasive impact on the attitudes and behavior of consumers relative to manufacturer information and traditional advertising (Cheong & Morrison, 2008). Among researchers and practitioners, there is considerable debate about the relative impact of positive versus negative reviews. General insights in psychology (see Baumeister, Bratslavsky, Finkenauer & Vohs, 2001 for a review) and consumer research (Mizerski, 1982; Basuroy, Chatterjee & Ravid, 2003) would predict that negative information has a greater impact than positive information. On the other hand, research by East, Hammond and Lomax (2008) has shown that positive word of mouth is more impactful with regard to customers’ brand purchase probability than negative word of mouth. Similarly, Vermeulen and Seegers (2009) have found that positive hotel reviews have a significant positive impact on consumers’ attitude toward the hotel, while negative reviews have little impact.
220
221
ETMAAL 2013
BOOK OF ABSTRACTS
To obtain a better understanding of the apparent discrepancy, we examine the relationship between valence and credibility of word of mouth. While research in psychology (Baumeister et al., 2001) has shown that negative information is usually more diagnostic and informative than positive information, we argue that – in the absence of other information - positive reviews may be more credible than negative ones. We think that the lack of personal knowledge about the motivation of the unknown reviewer (who is in fact no more than a stranger) in combination with the previous finding that negative information leads to greater scrutiny (Ahluwalia, 2002) and a smaller impact on purchase probability (East et al., 2008), makes customers perceive positive reviews as more credible than negative ones. The current study examines whether the valence of a review determines the perceived credibility of the review and that this perceived credibility in turn influences the consumers’ attitudes toward a product (in this case a – fictitious – restaurant). In other words, we hypothesize the following model: [Note: The model can be consulted in the PDF-file] To test our model, we used a 2 (positive versus negative review) x 2 (abstract versus concrete review) between subjects experimental design. Participants were 195 bachelor students of a Flemish University College. The sample consisted of 121 men (62.1%) and 74 women (37.9%), who were between 19 and 32 years old (M = 19.22, SD = 1.81). Participants indicated that they regularly visit a restaurant (M = 4.41, SD = 1.37). Participants in the positive (negative) concrete condition read a positive (negative) concrete review about a fictive restaurant, while participants in the positive (negative) abstract condition read a positive (negative) abstract review. After they read the review, they were asked different questions to determine their attitude toward the restaurant, the reviewer and the review they have read. To test whether our proposed model is correct and the review credibility is, in other words, a mediator of the relationship between the valence of the given review and participant’s attitude toward the product, we used the Bootstrap method of Preacher and Hayes (2004). The valence of the given review was a significant (p < .05) predictor of the attitude toward the restaurant. When the valence of the given review was positive the participant had a more positive attitude toward the restaurant than when the valence of the given review was negative. Further, the valence of the given review has a significant (p < .05) indirect effect on the attitude toward the review and this through review credibility. The more positive the given review was, the more credible participants’ perceived the review, in turn resulting in more positive attitudes toward the restaurant. This indirect effect partially mediated the relation between the valence of the given review and the resulting attitude. Practical implications of these findings will be discussed in the paper.
SETTING THE RECORD STRAIGHT: ONLINE REVIEW INTENTIONS AS A FUNCTION OF DINING EXPERIENCE AND PRIOR EXPECTATIONS - JEANA FROST, IVAR VERMEULEN AND ROMY VAN DER LINDEN What determines whether a customer engages in electronic Word of Mouth (eWOM)? In studies of traditional WOM, findings are mixed with some studies indicating that satisfied customers are more likely to engage in WOM while other work has failed to find a relationship between satisfaction and WOM intentions. To resolve this discrepancy, we suggest that eWOM is shaped, not simply by the 221
222
ETMAAL 2013
BOOK OF ABSTRACTS
level of satisfaction, but also by the expectations that preceded it. Both the extremity of the experience and its incongruence with prior expectations may prompt, in this case, eWOM. Consistent with our predictions, an empirical study finds that both positive and negative experiences yield higher review intentions than neutral experiences and incongruent experiences result in greater review intentions than mildly incongruent or congruent ones. Collections of reviews, often taken at face value, may in fact exhibit systematic bias.
DE INVLOED VAN SOCIALE EN INTERPERSOONLIJKE IDENTIFICATIE OP DE BEOORDELING VAN ONLINE REVIEWS - ALEXANDER SCHOUTEN AND MARJOLIJN ANTHEUNIS Online reviews worden steeds meer gebruikt in het aankoopproces. Een van de kenmerken waarop online reviews verschillen van traditionele vormen van word-of-mouth is dat de reviewer vaak onbekend is, waardoor de betrouwbaarheid van de reviewer moeilijker is in te schatten. Mede hierom geven online review websites steeds vaker informatie die klanten kunnen helpen een mening te vormen over de review en de reviewer. Zo gebruikt booking.com brede categorieën van reviewers zoals ouder echtpaar of jonge stellen en koppelt Tripadvisor reviews aan Facebook zodat ontvangers kunnen zien welke reviews hun eigen netwerk heeft geschreven. Er is nog weinig onderzoek gedaan naar de aard van de sociale relatie die een reviewer heeft met de ontvanger van een online review. Het doel van dit onderzoek is twee vormen van sociale relaties te vergelijken: sociale identificatie en interpersoonlijke identificatie. Sociale identificatie veronderstelt dat mensen meerdere sociale identiteiten hebben gelieerd aan bepaalde groepen waarmee men zich verbonden voelt. Bij elk van deze sociale identiteiten hoort een bepaald verwachtingspatroon van de normen, voorkeuren en gebruiken van deze groep. Toegepast op online reviews betekent dit dat als men weet dat de review afkomstig is van een lid bepaalde groep waarmee men zich identificeert, men deze review betrouwbaarder acht dan wanneer de review afkomstig is van een reviewer waarmee men zich niet identificeert. Interpersoonlijke identificatie is gebaseerd op de interpersoonlijke relatie tussen de reviewer en de ontvanger. Zoals word-of-mouth onderzoek al vele malen heeft aangetoond stellen mensen meer vertrouwen in de mening van bekenden dan van onbekenden. Bovendien blijkt de sterkte van de band ook een grote invloed te hebben op de betrouwbaarheid van een reviewer. Strong ties (hechte vrienden) worden over het algemeen meer vertrouwd dan weak ties (kennissen). Deze twee vormen van identificatie zijn nog niet vaak samen onderzocht. Daarom onderzoeken wij de rol van beide vormen op de geloofwaardigheid van online reviews, de attitude ten opzichte van het product, en de uiteindelijke koopintentie. De verwachting is dat reviews gemaakt door een iemand van een zelfde sociale groep betrouwbaarder zijn en een sterker effect op attitude en koopintentie hebben dan reviews gemaakt door iemand van een andere sociale groep. Daarnaast verwachten we dat reviews gemaakt door bekenden een groter effect hebben dan reviews van onbekenden, en dat dit effect sterker wordt naarmate de band met de reviewer sterker is. 404 proefpersonen (gemiddelde leeftijd 22 jaar, 91% student, 76% vrouw) namen deel aan een experiment waarbij de participanten verschillende hotelreviews moesten beoordelen. 256 proefpersonen kregen een review te zien die afkomstig was van ofwel een reviewer in dezelfde 222
223
ETMAAL 2013
BOOK OF ABSTRACTS
sociale groep (student) ofwel een ander sociale groep (ouder echtpaar). De overige proefpersonen kregen een review te zien die afkomstig was van een van hun Facebook-vrienden. Hiervoor zijn 22 personen gevraagd om hun Facebook ter beschikking te stellen en de hotelreviews onder hun naam door te sturen aan een willekeurige selectie van vrienden. Op basis van de band die de proefpersonen hadden met deze Facebook-vrienden werden de proefpersonen ingedeeld in strong & weak ties. Proefpersonen beoordeelden de reviews op betrouwbaarheid. Tevens werden attitude ten opzichte van het hotel en koopintentie gemeten. De resultaten lieten zien dat reviews van een andere sociale groep als minder betrouwbaar werden gezien dan reviews van een zelfde sociale groep, met als gevolg een negatievere attitude en lagere koopintentie. Tevens lieten de resultaten zien dat reviews van bekenden betrouwbaarder waren dan reviews van onbekenden uit een andere sociale groep, maar niet meer dan reviews van onbekenden uit een dezelfde sociale groep. Een verrassend resultaat was dat attitude en koopintentie negatiever werden beoordeeld wanneer de review afkomstig was van bekenden dan wanneer de review was gemaakt door een onbekende van dezelfde sociale groep. Dit resultaat gold zowel voor strong als weak ties. Blijkbaar hecht men meer waarde aan reviews van een onbekende binnen dezelfde sociale groep dan van bekenden. Een mogelijke verklaring hiervoor is dat men onbekenden ziet als stereotype deelnemers van een bepaalde sociale groep waarmee men zich identificeert. Hierdoor voelt men zich gelijk aan deze persoon en denkt men dat deze persoon ook dus dezelfde opvattingen heeft. Alhoewel men ook veel waarde hecht aan reviews van bekenden, heeft men over deze personen veelal meer informatie beschikbaar en kan men dus inschattingen maken van de kennis die deze specifieke reviewer heeft over het product. Een belangrijke conclusie is dan ook dat review websites niet blind moeten varen op een zo sterk mogelijk connectie met de reviewer, maar dat sterke effecten juist veroorzaakt kunnen worden door banden op basis van sociale in plaats van interpersoonlijke identificatie.
THIS T-SHIRT IS A-W-E-S-O-M-E!: THE CREATION OF CAMP PRODUCTS THROUGH IRONIC ONLINE REVIEWS - CAMIEL BEUKEBOOM, CHRISTIAN BURGERS AND AMBER SCHOLTZ On consumer web sites like Amazon, seemingly appalling items can unexpectedly rise like a star. The so-called “Three-Wolf-Moon-T-Shirt”, for instance, received thousands of jubilant reviews, that are read and found helpful by even more readers, and has a rating of 4.5 out 5 stars. Consequently, and despite its ugly appearance, the shirt has become one of Amazon’s top-selling apparel items (Applebome, 2009; Emery, 2009). The type of reviews that such products receive may explain this phenomenon. Reviewers mostly ironically describe how the owner of the shirt becomes irresistible to women (e.g., “The effect that this T-shirt has on women is pretty impressive.”) or obtains super powers (e.g., “Since owning the Three-Wolf-Moon-T-Shirt, I have successfully solved 7 crimes in my city, including 4 cold case murders.”). Other seemingly mundane products (e.g., “Tuscan Whole Milk”) have attracted similar types of reviews (e.g., "Has anyone else tried pouring this stuff over dry cereal? A-W-E-S-O-M-E!"). Apparently, people see something appealing in these products that invite them to write these extraordinary reviews. This phenomenon can be related to the broader concept of camp. When 223
224
ETMAAL 2013
BOOK OF ABSTRACTS
something becomes camp, people ironically embrace and glorify it in an undue manner. It is considered outlandish or corny, yet, and because of that, simultaneously perceived as something beautiful (Meyer, 1994; Sontag, 1964). This joint meaning is inherent to camp and is reflected in its definition “It’s beautiful, because it’s awful” (Sontag, 1964). This double meaning is exactly what irony can communicate. Reviews about products with camp appeal appear to recommend the product, yet exaggerate their praise to communicate an ironic opposite meaning. Senders alert readers that a message should be taken as irony, by using irony markers in the text (Attardo, 2000; Burgers et al., 2012). In reviews, ironic meanings can be communicated by exaggerating the intensity of evaluation (e.g., hyperboles; milk reviewed as “A-W-E-S-O-M-E!”). Reviewers can also exaggerate the review content, for instance by describing an outrageous unrealistic product experience, like attracting dozens of women and solving crimes wearing a t-shirt. This too can be regarded as irony, because an opposite meaning is intended (i.e., women will not be attracted and you will be anything but a superhero wearing this shirt). The present research focuses on how products obtain camp appeal in online consumer platforms as a result of the reviews they attract. We aim to identify what these reviews are like, and how they are different from regular product reviews. Moreover, we investigate the factors that make people write such reviews. In study 1, we content-analyzed Amazon reviews of actual products (N = 968 reviews, M = 104 words), and compared the reviews of products with apparent camp status to “normal” products from the same category. Results revealed a number of features that distinguish camp products and their reviews from other products. Where normal products were simply described and evaluated, reviews of camp products were longer and contained significantly more descriptions of unrealistic product experiences. Moreover, we observed that camp products contained a negative, unexpected aspect (either in the product itself, price, or description) that appeared to elicit ironic reviews. It also seemed that prior ironic reviews invited others to also write ironic reviews. In study 2 (N=143), we experimentally manipulated two factors in a 2 (Type of product: unusual vs. normal) x 3 (Presence of prior review: positive vs. negative vs. no review) between-participants design, to test whether these have the potential to elicit ironic reviews. Participants wrote a review on an Amazon page about either an unusual, unexpected product (huge, expensive Swiss knife) or a comparable normal product (regular Swiss knife). The page also showed either an extremely positive, negative, or no prior review of the product. Results confirmed that the unusual product, compared to the normal product, elicited significantly more irony. This was both apparent in participants’ hindsight judgment of their review, and in the coded actual content of the reviews. In line with the findings of study 1, reviews of the unusual product contained more descriptions of unrealistic product experiences than the normal product. The presence of a prior (ironic) review had only a minor effect on participants’ irony use in this experiment. In sum, our results suggest that unexpected, unusual products have the potential to elicit ironic reviews, which are distinguished from other reviews in that they ironically describe unrealistic
224
225
ETMAAL 2013
BOOK OF ABSTRACTS
product experiences, rather than describe and evaluate the product itself. Effects of obtaining camp appeal as a result of these reviews on product attitude and purchase behavior will be discussed.
Session #10 Friday 8 February 11:45 – 12:45 hrs. Chair: Joost Verhoeven Location: Rochester (M2-10) Dutch session
EVOLUTIONAIRE ROLLEN EN COMMERCIËLE COMMUNICATIE NAAR OUDEREN - DIRK SIKKEL In veel reclame wordt ingespeeld op de natuurlijke ontwikkelingsstadia van de mens, zoals die in de evolutie gestalte hebben gekregen. Emoties van kinderen, pubers, jong volwassenen, kostwinners en jonge moeders, gericht op het in stand houden van de menselijke soort, worden moeiteloos benut. Bij ouderen is dat anders. Over de opdracht die de natuur aan vijftigplussers heeft meegegeven bestaat dan ook veel minder duidelijkheid. Om te kunnen ontstaan en zich te handhaven moeten vijftigplussers een positieve bijdrage hebben geleverd aan de verspreiding van hun genen anders dan door voortplanting. Antropologen zijn het niet eens over welke deze bijdrage is. Er zijn verschillende hypothesen, onder andere (1) de omahypothese: oma’s hebben positief effect op de overlevingskans van de kinderen van hun dochter; (2) de moederhypothese: moeders zijn van een bepaalde leeftijd effectiever als ze hun reeds bestaande kinderen ondersteunen dan wanneer ze nieuwe kinderen baren; (3) de patriarchhypothese: sterke oude mannen verwekken kinderen die relatief oud kunnen worden en (4) de human resource hypothese: met het intelligenter worden werd ook het relatief lang vasthouden van kennis en ervaring effectief. Wat hier wordt beschreven zijn eigenlijk geen hypothesen, maar mechanismen die tot het bestaan van ouderen kunnen hebben geleid. Het is denkbaar dat niet één, maar meerdere mechanismen een rol hebben gespeeld. In onze samenleving vertalen de mechanismen zich als rollen. Naar analogie van advertenties voor jongeren is hier de onderzoekvraag: leidt herkenning van één of meer (potentiële) evolutionair bepaalde rollen van ouderen tot een hogere waardering voor een advertentie? En zijn dit dan ook de rollen die men zelf graag in het leven speelt? Om deze vragen te onderzoeken zijn in een internetsurvey onder vijftigplussers aan de respondenten 24 advertenties voorgelegd waarin een beeld van tenminste één vijftigplusser voorkwam. Er werd gevraagd om de advertentie te waarderen en de afgebeelde personen te associëren met een aantal rollen. De meeste hiervan waren afgeleid van hypothesen over de evolutie (bijvoorbeeld aanvoerder, verzorger, expert), maar ook enkele andere stereotypen zijn opgenomen (tobber, genieter). Daarnaast zijn vragen gesteld over de rollen die men speelt en wil spelen. Belangrijkste uitkomst is dat de match tussen rollen die men wil spelen en die men herkent inderdaad tot een hogere waardering leidt. De attributen van de advertenties konden worden teruggebracht tot twee beoordelingsdimensies: ‘verantwoordelijk’ en ‘gezellig’. De multiple correlatie tussen deze factoren en het rapportcijfer van de advertentie was 0.741. Ze fungeren als mediator tussen rapportcijfer en de geassocieerde rollen. De rollen ‘kostwinner’ en ‘expert’ beïnvloedden het cijfer via ‘verantwoordelijk’; de rollen ‘levensgenieter’, ‘verteller’, en ‘kenner van 225
226
ETMAAL 2013
BOOK OF ABSTRACTS
vroeger’ beïnvloedden het cijfer via ‘gezellig’. De rollen ‘adviseur’, ‘opvoeder’, ‘verzorger’ en ‘verantwoordelijke’ beïnvloedden het cijfer via beide mediators. Met name de rol ‘verantwoordelijke’ was de meest populaire en had ook de hoogste directe correlatie met de waardering van de advertentie. De uitkomsten van het onderzoek zijn in lijn met de hypothese dat inspelen op emoties die zijn gekoppeld aan evolutionaire rollen tot meer effectieve communicatie leidt, al vormen ze geen keihard bewijs. Het lijkt wel de moeite waard om deze benadering in de praktijk beter te exploreren.
NAAR EEN CONTEXTUEEL BEGRIP VAN DIGITALE UITSLUITING: DE ROL VAN SOCIAAL KAPITAAL IN ICT-TOEGANG EN –GEBRUIK VAN VROUWEN VAN MIDDELBARE LEEFTIJD - DANA SCHURMANS AND ILSE MARIEN Huidig onderzoek naar digitale uitsluiting erkent naast de invloed van socio-demografische variabelen zoals leeftijd, gender, inkomen, arbeidssituatie, familiale situatie, opleiding, etnische afkomst, taal en geografische locatie, ook de invloed van niet-socio-demografische variabelen zoals het sociale netwerk of de digitale hulpbronnen waarover een individu beschikt. Het blijft echter onduidelijk hoe en in welke mate het persoonlijk sociaal kapitaal een invloed heeft op de toegang tot en het gebruik van ICT. Sociaal kapitaal is een complex en multi-dimensioneel concept dat zowel verwijst naar sociale netwerken, sociale bronnen als naar vertrouwensrelaties en wederkerigheid. Bovendien is sociaal kapitaal een dynamisch begrip afhankelijk van de situatie (locatie) en de interpersoonlijke relaties (netwerken). Werk, familie, vriendenkring netwerken en netwerkrelaties zijn de voornaamste pijlers waarrond sociaal kapitaal wordt uitgebreid. Vrouwen van middelbare leeftijd worden binnen de huishoudelijke context traditioneel als ‘zwakke’ gebruikers geïdentificeerd omdat zij enerzijds niet tot de ‘digitale natives’-generatie behoren zoals hun kinderen. Anderzijds worden zij geconfronteerd met een klassieke paternalistische houding die zich ook vertaalt naar belemmering in ICT-gebruik –en toegang. Zij zijn minder in de mogelijkheid ICT te gebruiken voor het verbeteren van hun eigen sociale positie. Dit onderzoek focust daarom op de relatie tussen sociaal kapitaal en digitale uitsluiting en plaatst de volgende vragen centraal : Wat is sociaal kapitaal? Wat is de relatie tussen sociaal kapitaal en ICT-gebruik bij vrouwen van middelbare leeftijd? En welke rol spelen zijzelf binnen zogenaamd ‘ICT-sociaal’ kapitaal? Dit exploratief onderzoek vertrekt van een sociaal constructivistische invalshoek die de wisselwerking erkent tussen het effectief gebruiken en vormgeven van mediapraktijken. De benadering is theoretisch en empirisch. In eerste instantie wordt het begrip sociaal kapitaal theoretisch ontleed en vervolgens in het bredere veld van digitale uitsluiting geplaatst. Voor het empirisch onderzoek werden respondenten verzameld via een sneeuwbalmethode. In totaal vonden negen diepteinterviews plaats met vrouwen tussen 45 en 65 jaar. Aan de hand van proxytechnieken, met name verschillende stellingen en hypotheses, werden de vrouwen bevraagd over de rol van hun persoonlijke sociaal kapitaal bij het verkrijgen van toegang tot en het gebruiken van ICT. De literatuur insinueert een relatie tussen sociaal kapitaal en ICT-toegang en -gebruik maar de bijdrage blijft eerder oppervlakkig en beperkt tot de huishoudelijke context. De rol van sociale netwerken op (huidige) ICT-toegang en –gebruik wordt door de geïnterviewde vrouwen 226
227
ETMAAL 2013
BOOK OF ABSTRACTS
geminimaliseerd en ICT wordt eerder als een individueel iets ervaren. Echter, de impliciete bijdrage van elk netwerk en elke actor komt eveneens duidelijk naar voor in de diepte-interviews. Ieder netwerk, bron of wederkerigheidsrelatie omvat een specifiek soort kapitaal (technische ondersteuning, advies, online gesprekspartners,…) en vormt als het ware een ‘ICT-kapitaal’ dat bijdraagt tot het verkrijgen van toegang, het ontwikkelen van vaardigheden of een diversificatie van het gebruik. Bovendien is er sprake van een (impliciete) sociale druk om digitaal mee te zijn. ICT wordt voornamelijk vanuit professionele overwegingen geadopteerd en vervolgens in huishoudelijke en vrijetijdscontext geïntegreerd. De invloed van sociaal kapitaal op ICT-gebruik en –toegang is complex vermits ook de wisselwerking met en tussen economisch, cultureel en politiek kapitaal in acht dient genomen te worden om digitale uitsluiting te begrijpen. Het geheel aan soorten kapitaal maakt dat individuen enerzijds zelfstandig met ICT kunnen omgaan en ICT-gerelateerde vragen kunnen beantwoorden van hun familie-, werk-, vrijetijdsnetwerken (‘opinion leader’). Anderzijds kunnen zij beroep doen op deze netwerken wanneer nodig (‘opinion follower’). Sociaal kapitaal is niettemin een stimulans voor ICT-adoptie bij vrouwen van middelbare leeftijd.
WEBCARE ALS ONLINE REPUTATIEMANAGEMENT: HET GEBRUIK VAN WEBCARESTRATEGIEËN EN CONVERSATIONAL HUMAN VOICE IN NEDERLAND, EN DE EFFECTEN HIERVAN OP REPUTATIE JUDITH HUIBERS AND JOOST VERHOEVEN Sociale media bieden organisaties kansen om de relaties met stakeholders te verbeteren, maar communicatieprofessionals zien de open, ongecontroleerde dialoog op sociale media en het gemak van het verspreiden van informatie ook als een bedreiging voor de reputatie van organisaties (Verhoeven et al., 2012). Deze media zijn namelijk belangrijke fora voor het ontwikkelen van issues (Heath, 1998). Daarom zetten Nederlandse organisaties in toenemende mate webcare in om issues te monitoren en waar nodig in te grijpen. Door middel van een inhoudsanalyse en een experiment is onderzocht welke webcarestrategieën Nederlandse organisaties op Twitter gebruiken om reputatieschade te voorkomen, hoe effectief deze strategieën zijn en welke invloed een menselijke stem heeft op deze effectiviteit. Uit een inhoudsanalyse van 587 Twitterdialogen van 20 Nederlandse organisaties blijkt dat de informatiestrategie de meest gebruikte strategie is, gevolgd door sympathie en corrigerende actie. Opvallend was dat organisaties in veel gevallen meer dan één strategie tegelijkertijd gebruiken (40,6% van de webcaredialogen). Organisaties reageren soms met defensieve strategieën (rechtvaardiging en ontkenning), maar komen veel vaker tegemoet aan de individuele stakeholder (d.m.v. informatie, sympathie en compensatie), of aan de collectieve klantengroep (d.m.v. verontschuldiging en corrigerende actie). Uit een exploratieve factoranalyse bleek dat bovenstaande 3 clusters de werkelijkheid beter beschreven dan de gangbare tweedeling in offensieve en defensieve strategieën. Verder bleek dat organisaties de toon van berichten afstemmen op de inhoud: Collectief tegemoetkomende boodschappen werden geschreven met een meer menselijke toon dan niet-collectief tegemoetkomende boodschappen, en defensieve boodschappen werden geschreven met een minder menselijke toon dan niet-defensieve boodschappen. Ondanks dat de effectiviteit van boodschapstrategieën veelvuldig is onderzocht, weten we nagenoeg niets over de effectiviteit van webcare waarin verschillende strategieën tegelijkertijd worden gebruikt. 227
228
ETMAAL 2013
BOOK OF ABSTRACTS
Daarom hebben we vervolgens in een scenario experiment de effecten van de 3 clusters van webcarestrategieën (respectievelijk defensieve, individueel tegemoetkomende, en collectief tegemoetkomende strategieën) en van conversational human voice (wel versus geen menselijke stem) op corporate reputatie getoetst. Allereerst blijkt überhaupt dat het reageren op een klacht de reputatie kan verbeteren. Een individueel tegemoetkomende boodschap zorgt bovendien voor een significant betere reputatie dan een defensieve boodschap en een marginaal betere reputatie dan een collectief tegemoetkomende boodschap. De collectief tegemoetkomende strategieën en de defensieve strategieën verschilden niet significant van elkaar. Webcarestrategieën beschermen de reputatie dus vooral wanneer ze toegespitst worden op de individuele klant. Desondanks is het natuurlijk goed als organisaties klachten aangrijpen om structureel de dienstverlening te verbeteren. Klanten lijken vooral uit te zijn op de oplossing voor hun individuele problemen, en daarom doet een organisatie er goed aan om zich daarop te richten in webcare. Er werden geen hoofd- of interactie-effecten gevonden van de menselijke stem. Hoewel de manipulatiecheck bevestigde dat de manipulatie van de menselijke stem geslaagd was, liet deze analyse ook zien dat het verschil tussen beide condities niet groot was. Al met al bevestigen deze studies het belang van webcare ter bescherming van de reputatie van een organisatie. Organisaties beperken zich niet tot het gebruik van een enkele strategie, maar gebruiken strategieën synchroon. We leveren een bijdrage aan de conceptualisatie van webcare door het onderscheiden van 3 clusters strategieën. Organisaties beschermen hun reputatie het beste wanneer zij tegemoet komen aan de individuele belangen van de klagers.
DE KIP OF HET EI?: FACEBOOK & MERKERVARING - HARM TER HALLE AND JOOST VERHOEVEN Steeds meer organisaties investeren in een pagina op Facebook, zonder precies te weten welke impact dit heeft op de merkbeleving en merkrelatie. Het doel van dit onderzoek is om de impact aan te tonen van het volgen van een Facebookpagina van een merk op de merkbeleving, merkrelatie, merkbetrokkenheid, merkpercepties en koopintentie. Daarmee proberen we duidelijkheid te scheppen in de causale verwarring rond Facebookgedrag en merkrelatie: Versterkt het volgen van een Facebookpagina van een merk de merkrelatie? Of volgen mensen merken op Facebook omdat ze een goede relatie ervaren met dat merk? In lijn met de transportation theorie (Green, 2004) verwachten we dat narratieve structuren in een online merkgemeenschap op Facebook de Facebookgebruiker mee kan nemen in de wereld van het merk. Dit houdt in dat de ontvanger van een communicatieboodschap als het ware ‘ondergedompeld’ of getransporteerd wordt naar een narratieve wereld die gecreëerd wordt door het merk. Het publiek sluit zich bewust of onbewust af van de ‘echte’ wereld en laat zich meeslepen in het verhaal dat het merk communiceert. Facebook biedt merken de mogelijkheid deze verhalen te vertellen, en de gebruikers een positieve merkervaring te bieden. Omdat gebruikers door het volgen van merken op Facebook herhaaldelijk deze merkbeleving zullen ervaren (i.e., iedere keer als zij de branded content bekijken die op de timeline verschijnt), zal op 228
229
ETMAAL 2013
BOOK OF ABSTRACTS
termijn de relatie met dit merk versterken. We verwachten dan ook dat dit leidt tot meer merkbetrokkenheid, positievere merkpercepties en een hogere koopintentie. Om de hypothesen te toetsen is er gebruik gemaakt van een 2 (controle groep vs. experimentele groep) x 2 (voormeting vs. nameting) mixed factorial experimenteel design. De experimentele groep is gevraagd gedurende ongeveer 3 weken het merk Heineken te volgen via Facebook, terwijl de controle groep dat in dezelfde periode niet deed. Door het volgen van Heineken op Facebook worden de proefpersonen blootgesteld aan communicatie-uitingen van Heineken tijdens het Facebookgebruik. Er is voor Heineken gekozen omdat vrijwel iedere persoon het merk kent en omdat Heineken een actief Facebookbeleid voert. Gemiddeld zijn proefpersonen gedurende drie weken, in een natuurlijke setting, blootgesteld aan communicatie-uitingen van Heineken via Facebook in de maanden mei/juni. Aan dit veldexperiment namen 176 proefpersonen deel (waarvan 43% man, Mleeftijd = 28,6 jaar, SD = 5,2 jaar). Respondenten besteedden gemiddeld 55 minuten per dag aan Facebook (SD = 7,5 minuten). Repeated measures ANOVA’s laten significante interactie-effecten zien van het volgen van het merk Heineken en tijd (voor- versus nameting), hetgeen bewijst dat het volgen van de Facebookpagina van een merk de merkbeleving kan verbeteren, de relatie en betrokkenheid met het merk kan versterken en de koopintentie kan verhogen. Heineken is een zeer bekend merk, waar respondenten waarschijnlijk reeds vele jaren een relatie mee hebben. Via traditionele en nieuwe media, via sponsoring en andere kanalen worden respondenten met regelmaat geconfronteerd met marketinguitingen van Heineken. Toch toont deze studie aan dat zelfs het gedwongen volgen van een merk op Facebook in een relatief kort tijdbestek van drie weken significante effecten heeft op merkbeleving, merkrelatie, merkbetrokkenheid en zelfs koopintentie. Natuurlijk dient enige voorzichtigheid in acht te worden genomen bij de interpretatie van deze resultaten. Heineken besteedt bijzonder veel aandacht aan Facebook. Met grote frequentie wordt op Facebook vrij professionele content geplaatst. Juist dit kan onze resultaten verklaren. Vervolgonderzoek zal uit moeten wijzen welke effecten andere Facebook content heeft op gebruikers. Ook is vervolgonderzoek nodig om vast te stellen of het hier een kortdurend effect betreft, dat zich laat verklaren door de nieuwheid van de informatie op de timeline, of dat ook op langere termijn deze effecten blijven bestaan. Deze studie biedt een voorlopig antwoord op de causale vraag van de kip en het ei: Het volgen van een merk op Facebook heeft invloed op de merkbeleving en een merkrelatie. Dat neemt niet weg dat een relatie tot een merk gebruikers ook kan motiveren om een merk te volgen op Facebook. Daar merkrelatie zich moeilijk laat manipuleren, lijkt het lastig om door middel van een zuiver experiment met zekerheid vast te stellen of een merkrelatie ook leidt tot het volgen van een merk op Facebook. Pas als ook dit vraagstuk is opgelost, kan een definitief antwoord geven op de vraag van de kip en het ei.
229
230
ETMAAL 2013
BOOK OF ABSTRACTS
Dit onderzoek bevestigt de effectiviteit van Facebook als marketinginstrument. Door de inzet van Facebook zijn merken in staat nog meer mensen te binden aan het merk. Facebook verdient daarom een plaats in de marketingmix.
230
231
ETMAAL 2013
BOOK OF ABSTRACTS
POLITICAL COMMUNICATION Session #1 Thursday 7 February 14:05 – 15:05 hrs. Chair: Rens Vliegenthart Location: Heidelberg (M1-16) English session
PREDICTING PARTICIPATION: PERSONALITY, POLITICAL ATTITUDES AND THE USE OF PARTICIPATORY MEDIA - TOM P. BAKKER, KLAUS SCHOENBACH AND CLAES DE VREESE The ubiquitousness and popularity of participatory media like blogs, Twitter and social networks have led many to declare the age of ‘citizen participation’ (e.g., Benkler, 2006; Bruns, 2008; Shirky, 2008), with mainstream media losing their monopoly position as a public platform for political information and discussion (Coleman & Blumler, 2009). But while media (Domingo et al., 2008, Thurman, 2008) and politics (Francoli & Ward, 2008; Meraz, 2007) are increasingly focusing on participatory media, user base remains relatively small. This leads to the natural question: What constitutes this active group of people? In this study we propose a number of hypotheses based on theories and findings from communication and political science and neighboring disciplines, aiming at shedding light on the individual characteristics that help explain the political uses of participatory media such as Twitter, social networks and weblogs. More specifically, we examine if the personality trait characteristics extraversion and openness to experience are – in tandem with more conventional predictors – significant predictors of participatory media use. We hereby draw on the nascent literature that has demonstrated that individual personality traits can play a significant role in guiding politically related attitudes and behavior (Mondak, Hibbing, Canache, Seligson, & Anderson, 2010; Orchard & Fullwood, 2009). It has been found that personality traits have a genetic basis and are thus heritable (Van Gestel & Van Broeckhoven, 2003), and that they can explain variance in attitudes, opinions, and behavior used across various disciplines. Following improvements with regard to adequately, but also parsimoniously measuring personality, evidence has accumulated demonstrating that an individual’s personality – “a disposition to react in certain ways” (Blais & St-Vincent, 2011, p. 414) –affects various forms of politically related behavior (Gerber et al., 2011; Mondak & Halperin, 2008). We argue that the driving individual factors behind participatory behavior merits closer examination, given the notions that traditional forms of participation are in decline while online participatory platforms could serve as attractive alternatives (Gil de Zúñiga, Jung, & Valenzuela, 2012). Besides the role of personality, we take into account the role of political interest, talk and online news consumption by testing the ‘mediation hypothesis’ (Blais & St. Vincent, 2011; Gallego & Oberski, 2012), which proposes that the effects of personality on participatory behavior are mediated by conventional predictors of political participation. 231
232
ETMAAL 2013
BOOK OF ABSTRACTS
Based on original Dutch survey data (n=2,081) from December 2009, we find evidence that typical psychological variables as well as established political participation predictors help to explain the use of tools like Facebook, Twitter, forums and the like. Extravert and ‘open’ people are more likely to use these media. Moreover, their personality increases the chance that they talk to others about politics (given their preference for new and social situations) and also makes it more likely that they are interested in politics and consumes online news (because they tend to be outgoing, curious and have many interests), which in turn influences participatory media use. Further, the negative effect of political efficacy for some relationships, suggest that those who have a low to average confidence in the responsiveness of the political system are more likely to turn to alternative ways of engagement, such as contributing political content to participatory media. The explanatory power of personality variables advances our understanding of what political users of participatory media ‘are like’. Hitherto, there has been no or little research that has combined the insights from psychological studies and traditional political participation studies to explain the use of participatory media. While our study has only examined two psychological factors that could play a role, we contend that there are most likely more extant psychological mechanisms that may serve as handles for sketching a more complete profile of participatory people on the Web. This feeling is corroborated by the fact that our analysis included three typical strong predictors of political participatory behavior: talk, interest and news use (Boulianne, 2009). However, three of the four direct relationships between personality and using participatory media remained significant. As the relationship between personality and behavior is context-dependent, future research should focus on further disentangling the complex interactions between specific personality traits, social settings and media use – which together can impact how people think, feel or behave.
SPIRAL OF POLITICAL LEARNING. THE RECIPROCAL RELATIONSHIP OF NEWS MEDIA USE AND POLITICAL KNOWLEDGE - JUDITH MOELLER News media are essential for political learning. By covering events and developments in the political world news media contribute to the political knowledge of their audience (Graber, 1997; de Vreese & Boomgaarden, 2006). Political knowledge, on the other hand, is a powerful predictor of news use (Price & Zaller, 1993). Moreover, knowing about politics is a precondition to make sense of news stories (Eveland, Martin & Seo, 2004). Without basic political knowledge new information cannot be processed. News use and political knowledge are therefore mutually interdependent. From a longitudinal perspective the mutual interdependence means that the process is ought to be a selfreinforcing spiral. In this paper we analyze the dynamics of learning from news media in detail using data from a three wave panel survey study of 888 adolescents.
INTERNETGEBRUIK EN BETROKKENHEID BIJ DE POLITIEK - MARJON SCHOLS, JOS DE HAAN AND JEROEN JANSZ Al geruime tijd bestaat er een discussie over de groeiende afstand tussen burgers en de overheid in Westerse landen (Parent, Vandebeek & Gemino, 2005). Het gebruik van internet zou participatie en betrokkenheid kunnen faciliteren (Parent et al., 2005; Anduiza, Catijoch & Gallego, 2009), doordat online participatie relatief gemakkelijk, goedkoop en snel is (Polat, 2005). Door onder andere de 232
233
ETMAAL 2013
BOOK OF ABSTRACTS
grote hoeveelheid beschikbare informatie en online platforms is bijvoorbeeld het zoeken naar politieke informatie of deelnemen aan online discussies makkelijker geworden. Over sociale media (Twitter, Facebook) wordt vaak beweerd dat ze politieke participatie makkelijker maken, waarbij wordt gewezen op de activiteiten van zowel burgers als politici. Over de daadwerkelijke invloed van internetgebruik op participatie van burgers verschillen onderzoekers van mening (Bouliane, 2009). Zo zijn er onderzoekers die stellen dat internet politieke participatie makkelijker maakt en daarmee tot meer participatie leidt, terwijl anderen aangeven dat de online participerende mensen degenen zijn die offline ook al politiek actief waren (Anduiza et al., 2009; Norris, 2001). In eerder onderzoek lag de nadruk vaak op de invloed van internetgebruik op actieve vormen van participatie, zoals het deelnemen aan online discussies. Het internet heeft echter meer impact dan alleen op actieve politieke participatie. Zo komt men via sociale media ook ongevraagd politieke informatie tegen en gebruiken andere media zoals kranten en tv-journalen informatie verspreid via internet als input voor nieuwsberichten. In dit onderzoek naar de invloed van internet op de relatie tussen burgers en de politiek staat dan ook de bredere betrokkenheid van burgers bij de politiek centraal. Vinden burgers dat internet bijdraagt aan een grotere wederzijdse betrokkenheid van politici en van henzelf? Daarbij is het nog interessanter om te kijken hoe verschillende opvattingen hierover verklaard kunnen worden; waarom vinden sommige burgers dat het internet nu wel bijdraagt aan een grotere betrokkenheid bij de politiek en anderen niet? In dit onderzoek ligt de nadruk liggen op het verklaren van de verschillen in opvattingen. De centrale vraag van het onderzoek luidt dan ook: in welke mate vinden Nederlandse burgers dat internet bijdraagt aan de betrokkenheid van burgers bij de politiek en hoe kunnen verschillen in opvattingen hierover verklaard worden? Om de verschillen te kunnen verklaren worden de achtergrondkenmerken en factoren omtrent zowel individueel internetgebruik als politieke interesse onderzocht. Zo beschouwen politiek geïnteresseerde mensen de impact van internet mogelijk positiever dan mensen die niet betrokken zijn bij de politiek (Boulianne, 2009). Daarnaast beschouwen mensen voor wie het internet een belangrijke rol speelt in hun dagelijks leven dit medium mogelijk eerder als toegevoegde waarde voor betrokkenheid bij de politiek dan mensen voor wie het internet een minder grote rol speelt. De rol van internet wordt hierbij breed opgevat, zo wordt onder andere gekeken naar de mate van gebruik, het type internetactiviteiten en hoe belangrijk men internet vindt voor de samenleving. Het onderzoek is gebaseerd op de Nationale Internetenquête (SIDN, 2011), een online survey uitgevoerd in opdracht van Stichting Internet Domeinregistratie Nederland (SIDN) in het najaar van 2011. De online vragenlijst is ingevuld door 2366 respondenten en omvat vragen op o.a. het gebied van persoonlijk en zakelijk internetgebruik, normen en waarden, privacy, sociale relaties en politiek.
HOW DEBATES, THE NEWS AND VOTING ADVICE APPLICATIONS ADD TO ISSUE OWNERSHIP IN THE EXPLANATION OF SHIFTS IN PARTY PREFERENCES: THE 2012 DUTCH ELECTION CAMPAIGN - JAN KLEINNIJENHUIS, NEL RUIGROK, ANITA VAN HOOF AND WOUTER VAN ATTEVELDT Theory. Traditional theories of voting behavior that start from socio-economic background characteristics and ideology-based issue positions of voters have difficulties to explain huge shifts in party preferences in the short run of an election campaign. Empirical test of communication theories 233
234
ETMAAL 2013
BOOK OF ABSTRACTS
show that news about changes in the short run (e.g. stock exchange rates in the case of business news, successes in the polls and in debates in political news) exert an influence on preferences in the short run. A first question of this paper is whether a short-term influence of the news can be assessed indeed, but the ultimate research question is how the short-term influence of the news interacts with long-term ideology based perceptions of issue ownership. Method. A content analysis of newspapers and television news programs during the election campaign 2012 (July-September 2012) resulted in roughly forty thousand news statements about issue positions of parties, successes and failures (gains and losses) attributed to parties, support and criticism for parties, and real world developments with regard to Dutch society and the Dutch economy. A three-wave panel study of public opinion was used to monitor in each phase of the campaign party preferences of respondents as well as their media consumption choices (early July n=1805 (“first wave”), week before elections n=1544 (“pre-election wave”), after elections n=1344 (“post-election wave”). In the first wave of the panel survey issue ownership predispositions were assessed, and in the pre-election wave whether Voting Advice Applications (VAAs, like Stemwijzer and Kieskompas) were used and, if so, which advices were obtained. In addition respondents identified who did win the television debates. Which news was consumed in which phase of the campaign by which respondents was reconstructed by the attribution of news characteristics from newspapers and television news magazines to respondents based on their personal media consumption choices in each campaign phase. Preliminary results. The results show that issue ownership plays a major role in election campaigns, in case issue ownership associations are primed in later campaign news. In the first wave of the election campaign the issue ownership profiles of parties overlapped each other to a considerable degree (e.g. with respect to health care, employment, Europe). Huge shifts in party preferences in the pre-election wave can be observed for voters with ambiguous issue ownership associations who received unambiguous VAA-advices, or who perceived one of “their” parties as the winner of the television debates. The use of social media does not appear to affect party preferences directly, but a huge indirect intermedia agenda setting effect can be observed. The first assessment that Diederik Samsom, the leader of the Social-Democrats (PvdA), did win the first television debate – labeled as the “prime minister debate” – was based on a sentiment analysis of tweets during the debate on August 26th 2012. Nobody ever questioned afterwards whether Samsom did win indeed, and based on rising expectation he even was perceived as the winner of the next television debate. The August 26 debate resulted in an earthquake in political news in newspapers and television news magazines. After the debate the PvdA was portrayed as the winner, the PvdA-VVD-battle instead of the SP-PvdA battle was portrayed as the major cleavage in politics. The immigrant issue was further degraded. The follow-up news in newspapers and television did have considerable effects on voters.
234
235
ETMAAL 2013
BOOK OF ABSTRACTS
Session #2 Thursday 7 February 15:15 – 16:15 hrs. Chair: Linda Bos Location: Heidelberg (M1-16) English/Dutch session
MEDIA AND/ OR REALITY: MEDIATED INFORMATION, A REPLACEMENT FOR, OR AN ADDITION TO REAL WORLD DEVELOPMENTS AS A SOURCE OF ANTI-IMMIGRANT SENTIMENT IN THE NETHERLANDS, DENMARK, AND SWEDEN, 2003-2010? - MARIJN VAN KLINGEREN, HAJO G. BOOMGAARDEN, RENS VIEGENTHART AND CLAES DE VREESE Theory and hypotheses A great body of literature has studied the causes of anti-immigrant sentiments (Ceobanu & Escandell, 2010; Coenders, 2001; Schlueter & Davidov, 2011), as it affects ethnic inequality, induces segregation, and reduces cohesion in today’s Western societies (Schlueter & Davidov, 2011). Most of these studies have a threat-based approach: the realistic group conflict theory (Austin & Worchel, 1979; Levine & Campbell, 1972) states that competing groups (e.g., migrants) create competition of scarce resources, such as jobs and housing, cultural identity, values or even political power. This competition, or threat then leads to negative associations with the outgroup (e.g., immigrant) and more positive associations with the ingroup (e.g., natives). Some assume that an increasing number of immigrants directly affects anti-immigrant sentiment of citizens of the host countries (Scheepers, Gijsberts & Coenders, 2002; Semyonov, Raijman & Gorodzeisky, 2006). However, we argue that many of these studies do not explore how information about these immigrant groups reaches host country populations, and thereby they leave underexplored the role of media in the process of antiimmigrant sentiment dynamics. Research has shown that people are hardly ever aware of Real World Development (RWD), such as immigrant group size, yet they often do have a perception of it (Sides & Citrin, 2007). The coverage of immigration in the media barely represents the dynamics in the real world (Vliegenthart & Boomgaarden, 2007). Yet, people base their ideas on what they see in the media, and familiarize themselves with (incoming) ethnic groups through mediated information (Maio, Esses & Bell, 1994), rather than direct experiences and real contextual developments. Therefore, we investigate the impact of media salience against the role of RWD’s with regard to immigrants on the development of anti-immigrant sentiment. Hence, we investigate the following research question: To what extent do news media and Real World Developments play a role with regard to anti-immigrant sentiment? In this study we test whether: an increasing number of immigrants (H1), media salience (H2), and negative news reports (H3) will enhance anti-immigration sentiment among natives. Previous studies have found that “News effects can be moderated by the context in which messages are received” (Boomgaarden & Vliegenthart, 2009: p. 519). Hence, this research will be done in three West-European countries over a seven year time period. Meanwhile we control for key-events that took place during the time of observation, which could also affect anti-immigrant sentiments, such as the London and Madrid bombing, the Van Gogh homicide and the Mohammed cartoon. 235
236
ETMAAL 2013
BOOK OF ABSTRACTS
Data & method The study is conducted in the Netherlands, Sweden and Denmark, between the period 2003 and 2010. These three countries are compatible with regard to their economic situation (all in the top-20 of the OECD GDP rank list), political system (all multi-party systems), and media news outlets (a combination of public and private television broadcasters; quality and tabloid newspapers). For the analyses we needed three types of longitudinal data: Real World Data, which we collected from the Eurostat website; Media data, which are newspaper articles, from one tabloid and one quality newspaper in each country, about immigration. These were collected from online newspaper archives from each of these countries, and were systematically coded by native speakers; and Survey data, for which we used several waves of the Eurobarometer. The assumptions are tested using hierarchical multilevel regression analyses for each country separately, with time (i.e., half yearly figures) on the highest level and respondents nested within each period. Results Preliminary results over the three countries indicate the effect of immigration flows on immigrant attitudes to be largely mediated by media salience. Meanwhile, negatively toned media reports enhance anti-immigration sentiment. These results show most convincingly in the Netherlands and Denmark. The differences between the countries could be due to historical differences with respect to the immigration issue. Where it has been a dominant issue in Denmark and the Netherlands for some decades, it has only recently gained importance within the Swedish context (Green-Pedersen & Krogstrup, 2008).
MEDIA COVERAGE OF MUSLIMS: THE FOREIGN THREAT VERSUS THE DOMESTIC OTHER - ANOUK VAN DRUNEN Media inform citizens about social reality. Especially when issues or objects are unfamiliar to citizens, media’s perspectives can be very influential (e.g. Iyengar & Kinder, 2010; Mazzoleni & Schulz, 1999). To understand the effects that media may have, it is essential to investigate actual media content. This investigation focusses on the study of news print media in the Netherlands about Muslims. There are some scientific studies investigating how this topic is covered in the media (e.g. Foster, Cook, Barter-Godfrey & Furneaux, 2011; Nickels, Thomas, Hickman & Silvestri, 2011), however recent investigations of media coverage of Muslims are still low in number. Consequently an analysis of six months of print media coverage of Muslims is executed. It is investigated: (1) what the visibility of the topic is (2) what the visibility of certain frames is (3) how visibility of topics and frames differs for various outlets and (4) how visibility of topics and frames differs for domestic versus foreign news. A categorization of frames about Muslims was made based on a literature study, which resulted in four frames: a problematization frame, a fear and threat frame, a homogenization frame, and an otherness frame. Those frames were operationalized using a description of the “(…) central organizing idea, or frame, for making sense of relevant events, suggesting what is at issue.” (Gamson & Modigliani, 1989, p. 3).
236
237
ETMAAL 2013
BOOK OF ABSTRACTS
Furthermore a topic list was assembled based on pre-studies of the media content about Muslims. The visibility of the issue was measured with the amount of articles over time, the number of words spent and the number of times the word Muslim was used. Finally it was coded whether news appeared in the domestic or foreign sections of newspapers. To distinguish between outlets I use the amount of political versus entertainment information in newspapers (Bakker & Scholten, 2009) resulting in: quality, popular and free newspapers. The period of investigation was the 15th of September 2010 to the 24th of March 2011. The main topic most often discussed in relation to Muslims in print media is politics. After that respectively: violence/terror, Arabic spring, the PVV/Geert Wilders and experiencing religion are discussed most often. These topics reflect the current events of the time period. The uprisings in the middle-east were prominent as well as the trial against Geert Wilders. Regarding frames, the problematization and homogenization frames are most visible in coverage of Muslims. Next is the fear and threat frame and finally the otherness frame is least visible This shows that there mainly is a focus on Muslims as a problematic group of people, a group of people connected to criminality or represented in contrast to another group of individuals. Also, Muslims are usually portrayed as one group of people instead of as individuals. Both these frames are present in half of all coverage of Muslims, which is a considerable amount. The amount of attention for Muslims in outlets differs substantially. Quality newspapers pay most attention to the issue: 74.2 per cent of all articles published are in those outlets. Overall the findings show that the topic is most visible in quality newspapers. However, findings show no significant differences in the visibility of frames in the different outlets. These findings indicate that different types of outlets cover this issue in similar ways and that popular newspapers are not necessarily more focused on negativity, such as violence and threats (e.g. Moore et al., 2008; Richardson, 2009). Finally regarding differences between domestic and foreign news coverage of Muslims, the former spends more words on the topic but overall the issue is covered about equally in domestic and foreign news. The visibility of the four frames does differ significantly for domestic and foreign news The problematization and fear and threat frames are most visible in foreign news and the homogenization and otherness frames are most visible in domestic news. The findings indicate that especially ‘foreign Muslims’ are represented as dangerous and a threat. ‘Domestic Muslims’ are more often represented as a homogenous group and as different to an in-group, which shows that the media mainly see ‘our own Muslims’ as different but perhaps not necessarily as a threat.
GEBRUIK VAN SOCIALE MEDIA EN DEELNAME AAN DE EGYPTISCHE OPSTAND: DE ROL VAN TOEKOMSTVERWACHTINGEN, DEMOCRATISCHE ATTITUDES EN VERTROUWEN IN DE OVERHEID PETER KERKHOF, ANDRE KROUWEL, JACQUELIEN VAN STEKELENBURG AND BERT KLANDERMANS Previous studies have shown that social media use among participants in the 25 January Egyptian uprising was high and that protesters indicate that social media were important in their decision to join the protests (Tufekci & Wilson, 2012; Wilson & Dunn, 2011). However, other studies suggest that the attitudes of social media users toward the regime and toward democracy in general prior to the uprising were no different from non-users (Norris, 2012). So far, no direct comparison has been 237
238
ETMAAL 2013
BOOK OF ABSTRACTS
made between protesters and a equivalent group of non-protesters in terms of their social media use. In this paper, we use data that have been gathered by Vote Compass Egypt, an online voting tool, to compare protesters and non-protesters in how important they rate different sources of political information, among which social media. The results show that participants in the 25 January uprising rely more on Facebook and Twitter to get political information, and less on mass media (radio and TV). The association between protest participation and social media use is mediated by higher optimism about the future of the country, more democratic attitudes and less trust in government. Use of mass media is associated with less protest participation through more trust in government.
ARE TWITTER AND YOUTUBE THE NEW NETWORKED NEWS WIRES? A QUANTITATIVE CONTENT ANALYSIS OF JOURNALISTS’ SOURCING PRACTICES DURING THE ARAB SPRING - SARAH VAN LEUVEN, ANNELORE DEPREZ AND KARIN RAEYMAECKERS Since the 1980s concerns have been raised about the increasing commercialisation of the media. Numerous scholars state that the economic rationale prompts news organisations to stress costcutting and efficiency considerations. It is often argued that foreign coverage is one of the first victims of this ‘market-driven journalism’, for instance because news media no longer can or want to bear the cost of correspondents on the spot. As a consequence foreign news desks, for their information gathering, are increasingly depending on the international news wires and international news media whose information they often domesticate for their own national public. The result is a world view that is mainly dominated by official and powerful actors’ interpretation of events (Berglez, 2008; Buijs et al. 2009; Davies, 2008; Lewis, Williams & Franklin, 2008; McManus, 1994). The Internet and more specifically Web 2.0 applications or social media such as Facebook, Twitter or YouTube have yet re-opened the debate on the emancipatory potential of media use. In the context of globalization and digitalization, some scholars witness a shift from individualistic top-down monomedia journalism to participatory multimedia journalism characterised by networked communication, a synthesis of interpersonal and mass communication where audiences and journalists are connected in a networked media matrix. These authors state that the interactivity, connectivity, and flexibility of Web 2.0 applications invite a more diverse use of journalistic sources that might lead to a more balanced media access for a wider range of actors, including civil society organisations and individual citizens (Beckett, 2012; Brundidge, 2010; Castells, 2010; Dahlgren, 2005; Habermas, 2006; Heinrich, 2011). This ‘network’ or ‘networked’ journalism approaches world events from diverse – not only national and/or official – perspectives and thus enhances the quality of foreign news coverage. In practice several studies have found that journalists rarely incorporate social media in the news production process (Knight, 2011; Messner & South, 2011). Nonetheless we are hopeful as a few case studies indicate that journalists might be more inclined to ‘practice’ network journalism in the context of breaking news or media restrictions where they cannot (immediately) access the area themselves and therefore try to “learn from on-the-ground sources” (Lotan et al., 2011, p.1376). One recent example is the Arab Spring where Western journalists encountered difficult working conditions and thus may have been prompted to turn to their online and offline network of national and international informants. In the face of this assumption, we set up a quantitative content analysis of the prominent sources and actors in the Flemish news coverage of the Arab Spring. More 238
239
ETMAAL 2013
BOOK OF ABSTRACTS
specifically, we focus on three different cases - the revolts in Tunisia, Egypt and Syria – in four Flemish newspapers (two popular and two quality dailies) and two Flemish broadcasters (the public broadcaster VRT and the commercial channel VTM). The analysis shows that Flemish news coverage of the Arab Spring to a certain degree adapted to the new reality of ‘network journalism’. Compared to previous research (e.g. Van Leuven, Deprez & Raeymaeckers, 2012, in review), Flemish journalists more often displayed a global (in contrast to domesticated or national) perspective in their coverage, and more frequently consulted social media sources (especially Facebook, Twitter, YouTube and WikiLeaks). The tendency towards network journalism is also clear from the finding that individual citizens – not official actors such as politicians or government organs – are most frequently quoted. Furthermore, we see that the characteristics of network journalism are more prominently present in coverage of the Syrian uprising, which confirms that local circumstances (i.c. heavy media restrictions and absence of Belgian tourists or companies) are decisive in the implementation of network journalism.
Session #3 Thursday 7 February 16:45 – 17:45 hrs. Chair: Andreas Schuck Location: Heidelberg (M1-16) English session
QUALITY OF PUBLIC DEBATES - REGULA HÄNGGLI AND RICHARD VAN DER WURFF What is a good public debate? In this paper, I suggest that the quality of a public debate can be assessed based on two dimensions: communication style and range of views. Communication style is measured based on respect, civilism, and substance. Range of views is measured by the number of participants, the level of justification, and the wealth of views. The two dimensions allow the definition of Scandalous, Insider, Competitive, and Deliberative debates. In a final step, I explore influencing factors of both quality dimensions. The data stem from interviews with political actors collected in six Western European countries (France, Italy, United Kingdom, Denmark, Germany, Switzerland) in spring 2011. We interviewed them about debates on unemployment taking place from September 2010 to February 2011. The empirical analysis shows that the two dimensions are helpful in characterizing public debates and thus, assessing their quality.
IT’S FUNNY – BUT IS IT APPROPRIATE? POLITICAL HUMOR IN THE MEDIA AND ITS CONDITIONAL IMPACT ON CITIZENS’ SOCIAL TRUST AND EFFICACY - SOPHIE LECHELER AND ANDREAS SCHUCK Political humor in the form of sketches featuring parody or satire is increasingly popular which is why previous research has examined the effects of humorous media formats on citizens’ understanding of politics. However, these studies have focused almost exclusively on the US context (e.g., the Daily Show). Bringing in a European perspective, this study employs an experimental survey design (N=243) to test the effects of political humor (i.e. political satire in the media vs. traditional news) on citizens’ social trust and efficacy in the Netherlands. Participants in the experiment were exposed to video sequences stemming from a popular Dutch comedy format, Koefnon, and their reactions were 239
240
ETMAAL 2013
BOOK OF ABSTRACTS
tapped in a survey questionnaire in the lab. Previous research is ambiguous about the either positive or negative effects of political humor on audiences. Some emphasize positive effects on issue knowledge and attention for political topics among those least interested in politics. Others warn of the negative impact of political humor on political cynicism. As a result of these conflicting studies a debate has emerged discussing the normative implications of political entertainment in general and political humor in particular for society and democracy. Recent research has called for increased efforts to further disentangle the complex relationship between, e.g. different kinds of political humor (in terms of entertainment formats) on different kinds of outcomes. This is where the current study makes its contribution. Not only do we test the impact of political humor in a non-US context and focus on ‘parody’ as one distinct form of political entertainment particularly and increasingly popular in Europe – but we also introduce a new moderator for the presumed effects political humor has on citizens. We conceptualize this new concept as the individuals’ perception of political humor as an ‘appropriate’ ( vs. ‘inappropriate’) way to talk about politics. Previous research has engaged in testing the role of – by now – established concepts such as more general ‘need for humor’, i.e. the individual tendency of people to like and search for humor in their own personal lives, or more specific ‘affinity with political humor’, i.e. how much people like political comedy. The concept we introduce, ‘appropriateness of political humor’, is expected to be related but empirically distinct from these existing concepts, i.e. it is reasonable to assume that people who like and appreciate political humor can still differ with respect to their tolerance of what political humor may or may not make fun of. Vivid examples for this include the recent so-called ‘Vatileaks scandal’ and the cover title of the German satire magazine ‘Titanic’ showing pope Benedict XVI. wearing a white cassock which is soiled at crotch level. On the front cover, the stain is yellow; on the back cover, it's brown. The accompanying headline reads: "Hallelujah in the Vatican: The leak is found!" The Vatican sued the magazine and called the picture ‘intolerable’. Other illustrative examples include Angela Merkel wearing a Nazi uniform in a Polish satire magazine or the widely debated Mohamed caricatures which sparked violent demonstrations across the Arab world. Strikingly, public debate showed that commentators and common people alike disagreed on if such forms of political humor were ‘appropriate’ or not. The results of our study indeed show that this concept matters. First, watching a humorous TV sketch showed to result in lower social trust and external efficacy than exposure to a normal standard news clip on the same topic (i.e. in our study the Eurocrisis and situation in Greece). This negative effect, however, only occurs among citizens who have less affinity with political humor, and who consider it an inappropriate form of talking about politics. Interestingly affinity with political humor is particular important for the effect on social trust whereas appropriateness of political humor is particularly relevant with regard to external efficacy, again stressing the conceptual distinctiveness of both concepts. Thus, the results of our study not only demonstrate the relevance of our new concept of ‘appropriateness’ but also contribute to the more normative debate about the role of political humor in society. While we cannot say anything about potentially positive effects of political humor on issue knowledge, we can conclude that there are negative effects on social trust and external efficacy, i.e. the trust people have that politics is responding to what is important to them. However, the picture is not entirely pessimistic as our results also show that only those who have less affinity with such political entertainment formats and deem them an inappropriate way of talking about politics are negatively influenced.
240
241
ETMAAL 2013
BOOK OF ABSTRACTS
STYLE TALK. AN EXPERIMENTAL INVESTIGATION INTO DIRECT EFFECTS OF NON-TRADITIONAL POLITICAL COMMUNICATION STYLES ON THE PERCEIVED CREDIBILITY OF POLITICIANS - NIELS BOUWMAN AND LINDA BOS Political communication has entered an era in which responsiveness to the interests of ‘the people’ and the ways of the media seems prerequisite for electoral success. As a result non-traditional communication styles are on the rise. Political style and personality have conquered their spot next to substance. The building and keeping of political persona has become of key importance to a new breed of celebrity politicians for whom credibility is their greatest good (Corner, 2003; Street, 2004). Media proliferation has, at least for some, brought about the demystification of public office (Mudde, 2004). For the emancipated citizen politics is not, or should not be, a matter of hocus-pocus but rather a representative practice they expect to resonate with their interests for it to gain and maintain their electoral support. Under such conditions paternalistic discourse (a traditional asymmetric relation) is no longer an option to court popular support (Blumler & Kavanagh, 1999). Instead, a certain symbolic closeness (or symmetric relation) to the masses has become a necessary condition for electoral success (Garzia, 2011). Sensitivity for the interests of citizens and the ways of the media is believed to have resulted in a less elitist, more popular political communication (Blumler & Kavanagh, 1999). When political communication turns anti-elitist, it becomes populist; a widespread phenomenon in modern Western democracies according to Mudde’s (2004) populist zeitgeist thesis. Little has been done to empirically determine the individual-level influence of these non-traditional political styles on political judgment-forming. Answering to this deficiency and providing empirical substantiation to the ongoing debate on democratic merits and harms of politics’ popularization and a populist zeitgeist, this study investigates political communication style-effect by means of an online survey experiment among a sample of Dutch adult citizens (n=331). In doing so it connects two otherwise separate, yet interrelated, research fields; that of populism – studied mostly within the political sciences – and that of popularization – studied within media and cultural studies. A 3x2 between-subjects factorial experiment was designed to test the direct effects of a popular, populist and traditional communication style on the perceived credibility of both a mainstream and a populist fictive politician, quoted in a news article. The article quoted a fictive politician to prevent participants from having a preconceived image of the politician. The politician, named Bas Oord, was either member of the populist anti-immigrant party PVV or the mainstream central-liberal party D66. In explaining communication style-effects this study reviews the potential moderation effects of political cynicism, media entertainment preference, age, education and political leaning. All credibility’s components were first analyzed by means of an analysis of variance. Tests show significant mean differences across the six experimental conditions. Hierarchical linear regressions show, for both styles alike, a negative effect on perceived credibility of the politician, which is slightly stronger for populist style. Moreover, the party cue results in lower credibility ratings; the PVV politician is consistently judged less credible than the D66 politician. We do not find a significant interaction between party cue and style, which suggests that – overall – belonging to a populist party is no catalyst for further devaluation of non-traditional communication styles. We find no evidence that a populist politician can use a populist style without repercussions. Finally, we find a positive 241
242
ETMAAL 2013
BOOK OF ABSTRACTS
moderation effect of political cynicism on the negative populist-style effects. If party type is also taken into account we notice that the political cynical are less strongly (but still negatively) affected by populist style from a populist politician than the less cynical. Inversely, the more cynical seem more strongly and positively affected by a traditional style from a populist politician. Our findings show that people in general are not attracted to the populist or popular approach. Overall this seems to indicate that people do not feel better represented by its ordinary political communication styles. In fact, quite the opposite shows as a traditional – and in some way elitist – style considerably improved a populist politician’s credibility-ratings. This seems to indicate populist style is no cure; ‘normal’ politics simply has to step up its game.
THE PLAIN, FAMILIAR AND COMFORTABLE LANGUAGE OF THE COMMON PEOPLE. AN EXAMINATION OF RHETORICAL FIGURES IN POPULIST POLITICAL LANGUAGE - LIEUWE KALKHOVEN AND CHRIST'L DE LANDTSHEER It is often said that populism has made a revival in the last decades throughout Europe. This ambiguous phenomenon appears to be one of the most difficult political concepts to define and to fully understand. How exactly should populism be understood and what explains the success that many populist parties have – or at least temporally – in attracting voters? Is there something like the populist ideology, where people can hold on to, or does populism only exists by means of shallow and trivial rhetoric? Theories on populism often make a distinction between a populist ideology on one hand and a populist (communication) strategy on the other hand. Populism considered as an ideology can be described as the strive for ‘the will of the people’ and more direct democracy in order to win back the sovereign state from a corrupt and power hungry political, intellectual and cultural elite. Although this notion of populism shows a clear moral principle, it lacks defined and universal programmatic key ideas (unlike the usual universal ideologies, such as socialism or liberalism). The lack of fixed core (programmatic) ideas makes that the populist ideology is seen as a thin-centered ideology. Besides, substantively it is mainly based on a mythical and unreal conception of society and politics. The strength of populist parties derives certainly from the ability to address so many different voter groups, based on an attractive ideology (e.g. the idea of restoring the ‘power to the people’) and certain programmatic issues (e.g. ‘anti-immigration’ in many rightwing-populist parties’ programs), but in this study we argue that the populist success comes not at the least due to persuasive techniques, expressed by a strong charismatic leader. Already since ancient times people involved in politics are aware of the power of rhetoric, or the ‘art of speech and oration’. The political rhetoric describes language and style as a means to empower argumentation in order to persuade the public or political opposition in the debate. From a long tradition of research on persuasive political language it is known that rhetorical language above all aims at persuasion through emotions. Rhetorical style figures that appeal to emotions could therefore be effective tools in the political realm of public campaigning and debate. It is often argued 242
243
ETMAAL 2013
BOOK OF ABSTRACTS
that populist parties rely or even depend on rhetorical language, more than non-populist parties (or ‘mainstream’ parties). This study examines two rhetorical figures that are usually described as ‘typically’ populist: the particular use of metaphors and hyperbolism (or exaggeration). The two rhetorical style figures have in common that they operate on an emotive and latent level of political persuasion. A certain (ideological) version of reality is expressed figuratively and (therefore often) incorrectly. Whether this is represented in different words and meaning (through metaphors) or through an enlargement and overstatement (hyperboles), they all cause distorted and simplified versions of reality. By means of a comparative quantitative content analysis of official political parties’ utterances (parliamentary debates and press releases), the rhetorical language of populist parties versus mainstream parties in different time periods (2006-2011) and contexts (the Netherlands and Flanders) is being compared. As hypothesized, significant differences between several of these variables are detected. General results show that political parties appear to use more metaphors throughout the years. Moreover, parties’ role and position on the political spectrum are good predictors of rhetorical language: opposition parties and parties at the right of the spectrum use more hyperbolism in their language and are more negative in their overstatements. Parties at the center are the least hyperbolic and most positive when they use hyperboles, and show significantly less powerful metaphorical language. Parties that have been considered as populist parties have significant deviant scores than parties that have not been classified as populist, also on use of hyperboles, the value of hyperbolism and metaphor power. However, when the results of the individual parties are being considered, rather different scores are found. The results show nevertheless a defined distinction in rhetorical language between parties that have been considered populist and parties that are not.
Session #4 Thursday 7 February 16:45 – 17:45 hrs. Chair: Wouter van Atteveldt Location: Aberdeen (M3-03) English session
HOW POLITICAL CANDIDATES USE TWITTER AND THE IMPACT ON VOTES - SANNE KRUIKEMEIER, GUDA VAN NOORT, RENS VLIEGENTHART AND CLAES DE VREESE New media has become increasingly important during election campaigns. The potential of Internet to connect with and mobilize voters, gives politicians the opportunity to communicate with the electorate, without the interference of journalists (Blumler & Gurevitch, 2001; Golbeck, Grimes, & Rogers, 2010; Parmelee & Bichard, 2011). It is not surprising that political organizations have embraced the Internet. Political candidates are increasingly using new media, such as Twitter, Facebook and other online platforms, to promote themselves, and to communicate interactively with citizens (Golbeck et al., 2010).
243
244
ETMAAL 2013
BOOK OF ABSTRACTS
The rise of web campaigning has also been subject to a growing amount of scholarly inquiry (Gibson & McAllister, 2006). Although previous work gives valuable insights into both the content of online political communication and the potential effects, studies that provide a link between the two arenas are relatively scarce (Gibson & McAllister, 2006). Especially studies that examine the persuasive effect of different styles of online political communication have been limited. Prior work within this field has mainly concentrated on the effects of more general forms of web campaigning, for example Web presence of politicians and parties (Wagner & Gainous, 2009). However, politicians promote themselves differently online: some politicians communicate in an interactive way, while others are less interactive (e.g., reacting on tweets vs. posting information). Yet others exchange information about their private lives, while others do not. Little empirical studies have been focusing on these two important characteristics or styles of online campaigning (i.e., interactivity and political personalization) and we can infer even less about the consequences of these characteristics on voters. The current study tries to fill this gap by investigating the relationship between candidates’ style of online communication during an election campaign and electoral support for those candidates. This study focuses on the micro-blogging platform Twitter. Twitter has become an essential and frequently used medium during election campaigns – besides Web sites and Facebook (Parmelee & Bichard, 2011; Tumasjan, Sprenger, Sandner, & Welpe, 2010). However, little is known about the content characteristics of political tweets and even less is known about the impact Twitter has on voting. So, by focusing on candidates’ Twitter use, this study sheds new light on the link between the content of new media and its effects. In this study, three hypotheses are proposed. Based on previous studies that examined the effects of web campaigning on vote choice (e.g., Wagner & Gainous, 2009), it is expected that candidates’ use (vs. non-use) of Twitter is positively related to electoral support (H1). Based on the literature of interactivity (e.g., Lee & Shin, 2012) and political personalization (e.g., Rosenberg, Bohan, McCafferty, & Harris, 1986), it is expected that candidates, who use a more interactive or personalized style of campaigning, will receive more votes than those who were less interactive (H2) or personalized (H3). Two data sources are used to investigate the relationship: a quantitative content analyses of all tweets posted by political candidates during the three months leading up to the Dutch national elections on June 9, 2010, and two months after (N=40,957 tweets), and a aggregated dataset containing all ‘preferential votes’ and control variables for political candidates that were electable (N=445 political candidates). Based on preliminary results, it appeared that the use of Twitter by candidates increased significantly during the course of the election campaign (and significant decreased after the election day). As expected, Twitter is mainly used as a vehicle to talk about candidates’ private persona. Another important popular characteristic of campaigning on Twitter is the use of interactive features. Furthermore, hypothesis one is supported: candidate’s use of Twitter during the course of the campaign exerts a significant positive effect on electoral support (β=.100, p<.001), even when we controlled for the usual predictors that explain vote choice (e.g., position on the list and presence in
244
245
ETMAAL 2013
BOOK OF ABSTRACTS
traditional media). Hypothesis two is also supported: communicating in an interactive way has a significant positive impact on the amount of preferential votes a candidate receives (β=.096, p<.039). The results have several important implications. As the use of new media grows, the characteristics – and especially interactivity – become more important. Moreover, based on the outcomes of this study, it seems that using Twitter has positive consequences. Candidates who used Twitter during the course of the campaign received more votes than those who did not, and using Twitter in an interactive way had a positive impact as well. These findings answer a fundamental question about the effectiveness of online political campaigning: Using new media matters, especially when it is used in an interactive way.
EVELIEN D'HEER, PIETER VERDEGEM AND PETER MECHANT: TWEETING THE ELECTIONS? AN ANALYSIS OF TWITTER USE DURING THE 2012 LOCAL ELECTIONS IN FLANDERS Traditional mass media have long been the dominant medium for political communication. Within the last years, however, social media have entered the stage and are now perceived to support and facilitate more direct interaction between citizens and politicians. This way it is often argued that social media can function as a platform for political debate thus enabling a ‘new’ public sphere. Especially the micro-blogging platform Twitter seems to be given a prominent role in political communication during election times (Bruns et al., 2011; Larsson & Moe, 2012). This raises some important questions on the working and capacities that social media imply for citizenship and democracy. As mainstream and social media are part of the new hybrid media ecology, the interaction between them needs to be examined. This paper makes use of a case study of the Flemish local elections in 2012 with the goal (1) to shed light on the presence of political actors on Twitter during the campaign and (2) the Twitter conversation on a televised debate with the political actors. Using the Twitter Application Program Interface (API) we collected a corpus of 13.644 tweets using the most common hashtag of the local elections (#vk2012), between September 3th 2012 (the real kick-off of the campaign) and the election day (October 14th 2012). The televised debate with all Flemish political party heads, broadcasted by the Flemish public broadcaster VRT and held exactly one week before the elections, generated 2.615 tweets (#hetgrotedebat) and was watched live by 642.121 viewers. Twitter users, Twitter conventions and the content of the Twitter messages were subject of analysis. Based on our analysis we noticed only two political actors in the top ten of Twitter users on the local elections (#vk2012), namely the social democratic party sp.a (@sp_a) and a politician of the far-left PVDA (@TomDMeester). In addition, we found that the Green party (@groen), the social democratic party (@sp_a), and the liberal party Open VLD (@open_vld) are the top three users to which most @mentions were directed, which is a Twitter convention suggesting interaction, in this case with political parties. Closer investigation, however, shows that these tweets are mainly used for campaigning rather than facilitating a true dialogue. More specifically, the @mention is often used by the political parties themselves to refer to themselves. The currently most popular political party in Flanders, the nationalist N-VA (@de_NVA), has a rather modest activity in the political Twitter sphere. Nonetheless, they are in de top ten most retweeted users. In addition, the largest number of 245
246
ETMAAL 2013
BOOK OF ABSTRACTS
hashtags on political parties, were found on the N-VA as well: #NVA (N=452) and #N-VA (N=391). The high Twitter traffic on N-VA, in contrast to its low profile on Twitter, requires us to contextualize these findings in the broader media ecology, such as (the popularity of) broadcasting television (Vergeer in press). The televised debate on the public broadcaster VRT stages the political actors and their political interaction. The mass media catalase the formation of Twitter audiences (in our case using the hashtag #hetgrotedebat), coalescing into publics, engaging with conversation and interaction on politicians, their arguments, the topics under discussion as well as the media as an institute. During these televised debates, social media presence and activity of politicians is rather limited. Citizenviewers primarily reflect on the political actors, mainly as individuals rather than party members, and on their arguments. The expertise and credibility of the actor and the rationale of argument are discussed, aside aesthetic elements such as style, appearance and intonation. Hereby, audience members often use indirect forms of communication such as jokes, irony and metaphors. To conclude, we are careful in making grand claims about this development, as citizens using Twitter during TV programs are a vocal minority (compared to the number of viewers). In addition, we touch upon some prominent methodological, ethical and epistemological issues related to the collection, analysis and publication of big data.
HEIDI SCHULZE AND DAMIAN TRILLING: OF BINDERS AND BAYONETS: MEASURING THE ADOPTION OF FRAMES FROM THE US PRESIDENTIAL DEBATES ON TWITTER Since the first broadcast in 1948, the Presidential Debates have evolved as central events of each US election campaign, gathering high amounts of attention from the media and the interested public. While the earliest debates related to the presidential elections allowed the candidates to present their political programs vastly, today’s debates are highly regulated, media-oriented events. The candidates are forced to present their point of view and intentions as quickly and distinctly as possible, and hardly have time to confront each other directly (Trent & Friedenberg, 2008). Still, the debates offer the candidates the opportunity to present their political agenda directly to the public – and as it has been shown, to have direct and indirect influence on political attitudes and voter behavior (Pfau, 2002; Trent & Friedenberg 2008). Increasingly, Twitter is used simultaneously while watching television, acting as a ‘second screen’. This gives journalists, program makers, and researchers the possibility to gauge reactions of the public and public opinion instantly. Accordingly, a number of scholars have, for instance, tried to predict election outcomes using Twitter data (for an overview, see Gayo-Avello, 2012). However, tweeting about political events on television also adds a new layer of discourse to the debate, with potential effects of its own. To assess the impact of today’s Presidential Debates it is, therefore, crucial to know which aspects of the debates are picked up and disseminated by those active on Twitter. A fruitful framework for such an undertaking is the concept of framing. One can reasonably assume that within the setting of a Presidential Debate, politicians deliberately use specific framing techniques, referred to as strategic framing and applied by politicians to present and sell their view 246
247
ETMAAL 2013
BOOK OF ABSTRACTS
of an issue (Hallahan, 2008). To assess the role of Twitter as a second screen during these debates, our study addresses two specific questions: RQ 1: Which frames do the candidates use in the US 2012 Presidential Debates? RQ 2: To what extent is the adoption of these frames by Twitter users visible in their tweets? The construction of frames can be fairly well researched via content analyses, but frame adoption is harder to tap into. Experiments and linking opinion-poll results to long-term content analyses are commonly used to this end. Analyzing frame adoption by Twitter users, however, we are in the convenient situation that we can conduct a content analysis of their tweets in order to find out the extent to which frames used by the debaters are adopted by the audience. In order to answer the posed research questions, two content analyses are conducted. The aim of the first content analysis is to extract the frames presented by the candidates in each debate. Unit of analysis is therefore the transcripts of the debates. As the exact content of the debates is initially unknown and the overall context of a message is of central importance, a qualitative approach will be applied in order to analyze the promoted frames. The results of this analysis will be used as the basis for the second content analysis, which examines quantitatively the extent to which people adopt the presented frames. Unit of analysis are the tweets sent during the debate as well as the following night. To this end, on the days of the Presidential Debates, October 3, October 16 and October 22, all tweets including the keywords or hashtags of the debate, namely, debate, debates, debates2012 and PresidentialDebate2012, were recorded. Out of these, a random sample of N=10.000 tweets was selected. In addition to the occurrence of the frames in question, we analyze also the context in which they occur: For instance, the degree of interaction can be determined by looking at whether the frames occur in single tweets, in retweets, or in replies. Furthermore, the emotions and valence attached to the frames will be gauged. As the deadline of this abstract coincides with the week of the last Presidential Debate, the data have not yet been analyzed; however, this will be done before the deadline of the final paper. Results will provide valuable insight into frame adoption generally, but also into the role of Twitter as a ‘second screen’ discourse tool during political debates.
Session #5 Friday 8 February 09:00 – 10:00 hrs. Chair: Rosa van Santen Location: Heidelberg (M1-16) English/Dutch session
GAINING ATTENTION BUT COMMUNICATING LITTLE? ON THE CONDITIONALITY OF EPISODIC AND THEMATIC COVERAGE OF PROTEST EVENTS - RUUD JESSE WOUTERS Abstract –Social movements are underdog news sources that have to fight an uphill battle to make news. Protest actions are considered as one of the few avenues available for movements to gain 247
248
ETMAAL 2013
BOOK OF ABSTRACTS
media attention. The benefits of media attention for movements are manifold: media attention validates a movement, allows movements to reach out and seek social support. By diffusing their claims, media attention may enlarge the scope of a given conflict which, in a next step, can shift existing power balances to the advantage of movements. This article focuses on how protest is represented in the media arena. More specifically, whether coverage focuses on the details of the event (protestor behavior and appearances, the noise, spectacle and drama) or on the underlying issue that fueled the action, is subject of investigation. In the words of Iyengar (1991), this paper tests whether coverage of protest events is episodic or thematic, and teases out which factors drive thematic and episodic coverage. Extant research (Gitlin, 1980; Gamson & Modigliani, 1989: Smith et al 2001) leads us to expect that protest coverage is more episodic than thematic. Research on the conditionality of episodic versus thematic coverage is rare, yet different levels of gatekeeping theory (news values, news routines, media ownership) lead to specific hypotheses. Two data sources are used to test the above: protest event data gathered in Brussels police archives allows to construct a media independent picture of protest events. News coverage of the biggest commercial and public Belgian broadcasters is used to gather data on the media coverage of protest. Airtime devoted to the issue or the event within media reports on protest serves as dependent variable operationalization. The research period runs from 2003 to 2010; the number of unique protest events is 497, the number of protest reports is 650. Results show that protest coverage is more thematic than episodic. Most information in protest reports deals with the issue and its underlying conditions. Demonstration size is not a significant predictor of thematic/episodic coverage, yet whether the demonstration is disruptive (arrests, violence, property damage) is: more discordant protest leads to more episodic coverage. Variance explained by these real world conditions is rather low, however. If the event is staged by strong orgnazations that have higher news capability, coverage proves to be more thematic. These organizations can provide the press with information subsidies (Gandy, 1982) which allows them to gain leverage in the journalist-source relation. Especially protest staged by asylum seekers and migrants is covered more episodically. Protest staged by the more professionalized unions, environmental organizations and North-South organizations (development aid, human rights) is more thematic. This block of input variables, grounded in news routine research, explains the lion’s share of variance. Media ownership, finally, matters for protest coverage too. Protest on the public broadcaster is covered more thematically than on the commercial station. These results are discussed in light of extant American research, points to differences in media and political systems as foundations for divergent results, and proposes effect research as a promising avenue for further research on the importance of media coverage for movement impact.
MEDIA POPULISM OVER TIME - LINDA BOS AND KEES BRANTS The rise of populist parties in European democracies raises several questions. One of them, of particular importance to (Political) Communication Scientists, is how the success of these parties can be explained by media attention, but also, how media content is affected by right-wing populist success. In other words: an important research agenda in this regard is the relationship between political, or party, populism and media populism. 248
249
ETMAAL 2013
BOOK OF ABSTRACTS
Recently, a growing number of studies have looked into this matter. And even though questions have been answered, new ones can be raised. For instance, Bos, van der Brug and de Vreese (2010) concluded that the simplistic language, or populist style, used by populist party leaders strikes a chord with media logic. And their analysis of the Dutch 2006 election campaign also shows that the “alleged central populist tenets are not exclusively confined to (right-wing) populists” (Bos et al, 2010, p. 158). Connected to this is the alleged difference between the quality and the tabloid press in adopting media populism. A recent study by Tjitske Akkerman (2011) shows that popular newspapers do, in fact, not have an anti-elite bias. This corroborates evidence found by Bos et al. (2010). As said, these findings raise several questions. First of all, the chicken or the egg issue: what came first? Did the rise of (right-wing) populist parties affect the communication style of mainstream parties? Or were these parties already becoming more populist, maybe for other reasons? Is it a trend we are witnessing? And second, to what extent is this media attention for populist elements, such as a populist style or a populist ideology, really signalling a connection to media logic? Is it really a sign of media populism (Mazzoleni, 2003: 8), i.e., the “responsiveness to popular tastes and demands”? Or is more media attention for populism a sign of something else, maybe of a populist zeitgeist (Mudde, 2004)? That we do not find the expected differences between mainstream and tabloid media may also be taken as evidence for the end of a trend. Maybe there were differences between these media outlets more than a decade ago. All these questions can be answered by taking an over time perspective. In this paper we do so and whether there are signs of growing media populism in the media coverage of party leaders in Dutch election campaigns. In doing so, we distinguish between three aspects of media populism. First of all, we look at populist style elements. This style consists of various elements: dramatization, ordinariness, strong language, and strong leadership. The ‘empty shell’ of populist ideology includes anti-elitism and the celebration of the common man. Finally, we look at the right-wing populist ideology: being negative about certain sections of the population, and referring to the immigration and integration issue. The distinction between these aspects of media populism can learn us more about the degree in which the populist rhetoric is adopted by various mainstream and populist party leaders and by different media outlets. Moreover, we distinguish between controlled – party – publicity, and uncontrolled – mass media -- publicity. This setup allows us to investigate whether changes in mass media populism are propelled by changes in party communication strategies. We cover almost 20 years of Dutch election campaigns, and look at the campaigns for the general elections of 1994, 1998, 2002, 2003, 2006, 2010 and 2012: we use content analyses of the last four weeks before the elections. We include one broadsheet newspaper, NRC Handelsblad, and one tabloid newspaper, de Telegraaf, one public broadcast news program, NOS Journaal, and one commercial news program, RTL Nieuws. We also include several infotainment programs: Het Elfde Uur, de Grote Vier, Barend en van Dorp, Pauw en Witteman and Knevel en van den Brink. Finally, we look at Public Party Broadcasts (PBB’s) to investigate the amount of media populism in controlled publicity. The results will show the extent to which party leaders of various parties – mainstream and populist – are presented with a populist style, populist ideology, or even a right-wing populist ideology, in the controlled media as well as the uncontrolled media. By taking a longitudinal perspective we are able to answer questions, such as: what came first? Mainstream party leaders acting populist, or populist 249
250
ETMAAL 2013
BOOK OF ABSTRACTS
party leaders doing so? Do we see a difference in the adoption of ‘media populism’ by different media outlets over time?
DIRECTE AANDACHT VOOR RETORIEK EN NEGATIVITEIT - LUUK LAGERWERF, CHARLOTTE BRUSSEE AND AMBER BOEYNAEMS Politici strijden om aandacht van de media, en zetten daar onder meer hun toespraken voor in. Effectieve toespraken bevatten elementen die door journalisten als soundbites worden doorgegeven, of die applaus aan een welwillend publiek ontlokken (Bucy & Grabe, 2007; Bull & Wells, 2002; Kroon Lundell & Ekström, 2010). Speechschrijvers zetten frames en retorische middelen in om politieke en managementtoespraken aantrekkelijk te maken. Onderzoek naar onmiddellijke real time effecten van retorische middelen is niet vaak verricht. Toch is dergelijk onderzoek belangrijk want het kan verklaren hoe directe aandacht leidt tot een betere opslag van informatie en daarmee een betere herinnering (Potter, Lang, & Bolls, 2008). Het is ook niet eerder onderzocht hoe framing direct aandacht kan trekken (vgl. Fridkin & Kenney, 2004). De probleemstelling luidt: welke directe aandacht krijgen retorische formuleringen en valence frames in politieke toespraken? In het onderzoek werden twee toespraken over verschillende issues (onderwijs en ontwikkelingssamenwerking) gemanipuleerd, zodat per issue een positief en een negatief geframede variant werden gevormd. Deze varianten kregen elk een retorische en een niet-retorische versie: op zes prominente plekken werd in de retorische versie een retorische formuleringen aangebracht en in de andere versie een neutrale formulering. Als retorische formuleringen werd een deelverzameling gebruikt van in eerder onderzoek geïndentificeerde applaus genererende zinsconstructies, zoals contrast, herhaling-in-drie (drieslag), voortzetting of schelden (Atkinson, 1984; Bull & Wells, 2002). In een vooronderzoek onder 40 respondenten bleek dat respondenten meer aandacht hadden voor de negatieve speeches dan voor de positieve speeches. Aan de speeches met retorische formuleringen werd meer aandacht besteed dan aan de speeches zonder retorische formuleringen en deze werden ook beter onthouden. In een 2x2 design met herhaalde metingen van retorische formuleringen kregen 80 respondenten van dezelfde mannelijke stem verschillende varianten van de politieke toespraken te horen in een audiofile op een computer. Aandacht werd gemeten als secondary task reaction time. Respondenten keken naar het computerscherm dat oplichtte op de plekken waar (al dan niet) een retorische formulering werd gebruikt. Respondenten moesten dan zo snel mogelijk een toets indrukken. De reactietijden werden gemeten. Een langere reactietijd betekende meer aandacht voor het primaire proces van het luisteren naar de toespraak (Sternadori & Wise, 2010). Herinnering, begrip en waardering van de spreker en de boodschap werden post exposure gemeten. Positieve speeches zonder retorische formuleringen kregen minder directe aandacht dan andere speeches. Meer aandacht zorgde niet voor meer herinnering. De keuze voor een frame of de inzet van retorische formuleringen had verder geen impact op de waardering van de speech en de spreker, met uitzondering van de begrijpelijkheid van een speech. Een negatief geframede speech met retorische middelen werd beter begrepen dan de andere speeches. De resultaten bevestigen eerder onderzoek met andere toespraken en een andere wijze van meten van directe aandacht (op het Etmaal 2012 gepresenteerd).
250
251
ETMAAL 2013
BOOK OF ABSTRACTS
MEDIA CUES AND SUPPORT FOR RIGHT-WING POPULIST PARTIES - PENELOPE SHEETS, LINDA BOS AND HAJO BOOMGAARDEN Right-wing populist parties are thriving across Europe, currently holding seats in national parliaments in most European countries. Usual explanations for the success of these parties are based on notions related to economics, national identity or cultural alienation. Others, however, have argued that the mass media play a significant role in promoting the success of the populist right and have provided empirical evidence in support of this allegation (Walgrave and de Swert, 2004; Boomgaarden & Vliegenthart, 2007). Whereas those studies provide evidence drawing on macro-level data resources, little is known about how mass media coverage may affect support for the populist right on the level of the individual voter. Our study seeks to fill this gap. Populism is usually defined with regard to two features: anti-elitism and representation of the common people. The latter is often combined with strong national favoritism and hostility towards immigrants. Our experimental design draws on these two aspects of populist politics, incorporating both anti-politics and anti-immigration sentiments as media cues presented to respondents in a simulated news article. In addition, prior investigations often consider party cues to be important in the effects of media coverage on policy support. We thus add a cueing of the party of interest as a third experimental manipulation. Furthermore, to our knowledge all prior studies of support for populist right-wing parties have drawn on explicit measures of party support. We do also know, however, that survey measures of support for extreme right parties usually considerably underestimate the amount of support for these parties for reasons of social desirability. In this study we therefore employ two dependent variables to measure support for the populist right by means of both explicit and implicit measures. Accordingly, this study addresses both theoretical and empirical questions: First, to what extent does a message cueing immigration, anti-politics, or far-right parties boost support among voters for farright populist parties? Second, to what degree are implicit, unconscious measures of far-right support empirically consistent with explicit, self-report measures of vote likelihood? We used an experimental design embedded in an online survey of adult voters to address these questions. After a short pre-test measuring (among others) previous vote choice, political knowledge, political interest, and generalized trust, we randomly presented respondents with one of four versions of a short realistic news article with either an immigration cue (a minority group was mentioned), an anti-politics cue, a party cue, or no cue. A manipulation check showed significant differences between the four experimental conditions on recall of the cues. Respondents then took a Brief Implicit Association Test (Sriram & Greenwald, 2009), in which their implicit positivity towards a far-right party was captured. Afterwards they were questioned on their explicit attitudes towards that party, measuring their probability to vote, as well as other political attitudes and impressions. Results show higher implicit and explicit preference for the far-right party among respondents with high political knowledge, in response to the immigration cue. By contrast, low- and mediumknowledge voters were unaffected by any of the cues. Furthermore, only for high knowledge 251
252
ETMAAL 2013
BOOK OF ABSTRACTS
respondents were implicit and explicit attitudes correlated. This comports with literature suggesting the importance of knowledge for the effectiveness of media cues (Miller & Krosnick, 2000), but adds nuance to our understanding of the relationship between implicit and explicit political attitudes. Thus, our study expands our understanding of the effects media messages can have on far-right support, the relationships between explicit and implicit measures of far-right attitudes, and the differential effectiveness of competing political cues in this domain.
Session #6 Friday 8 February 10:15 – 11:15 hrs. Chair: Sophie Lecheler Location: Heidelberg (M1-16) English session
FRAMING CLIMATE CHANGE - EMOTIONAL RESPONSES AS MEDIATORS FOR THE EFFECT OF THREAT FRAMING ON INDIVIDUAL RISK PERCEPTIONS AND BEHAVIORAL INTENTIONS - ANDREAS SCHUCK AND LUKAS OTTO This study tests the impact of threat framing in news coverage about climate change and global warming on citizens' individual risk perceptions and behavioral intentions. Public debate on environmental issues such as climate change is widespread and increasingly intense, however, the question how the way media cover this debate affects citizens attitudes and perceptions surprisingly has not been thoroughly studied by previous research. This is even more striking considering that future policies regarding global climate change hinge on public support. The present study is one of the first to turn to the effect side of the issue. It employs a large-scale experimental design based on a representative sample (N=870) to test the impact of threat framing on how citizens think of the issue and puts special emphasis on the individual factors that moderate and mediate possible effects. More in particular, our research design takes three different factors into account. First, we distinguish between threat vs. no-threat frames. Previous research, for instance in health communication or regarding certain political issues like terrorism, has shown that threat frames can lead to more negative evaluations, increased participation rates and different emotional reactions. Importantly, these reactions depend on individual characteristics, such as pre-existing attitudes towards the issue or perceptions of self-efficacy, i.e. the perceived capability of coping with the threat. Previous studies suggest that individuals with high levels of self-efficacy tend to respond with ‘anger’ to confronting threats whereas individuals with low efficacy respond with ‘fear’ – both of these emotions are associated with different persuasive and mobilizing effects regarding subsequent attitudes and behavioral intentions. At this point we do not only take into account explicit, selfreported measures of emotion, but also apply the construct of implicit affect to communication research in order to come to a better understanding of unconscious affective processes underlying the described framing effects. Second, we distinguish between episodic vs. thematic news framing in our research design. Episodic frames present an issue from the individual perspective of a personal actor and/or based on a concrete event whereas thematic frames focus on facts and figures and put the issue in a broader context. Previous research suggests episodic framing to be more engaging for audiences and thus to evoke stronger emotional responses and, consequently, stronger effects on 252
253
ETMAAL 2013
BOOK OF ABSTRACTS
subsequent attitudes and behavior. Finally, the third factor in our research design consists of visual depictions of the topic. Since the stimulus material in our experiment consists of newspaper articles and given the nature of our topic (climate change) and our particular interest in the role of implicit and explicit emotions as mediators, we distinguish between threatening visual material (i.e. a photograph accompanying the text), neutral visual material and no visual material. Previous research suggests that the same news text is perceived differently depending on the arousing or descriptive nature of the accompanying visual material. However, our study design allows to test this relationship in interaction with the other two factors (threat framing vs. no-threat framing and episodic vs. thematic framing) which offers pathways for relevant follow-up questions, e.g. with regard to the relative importance of the framing compared to the visual material and the extent to which these can reinforce each other or cancel each other out. Preliminary findings indicate that threat framing does affect subsequent attitudes and behavioral intentions and that the effect is stronger in the episodic condition compared to the thematic condition. Furthermore, we focus on the moderated mediation dynamic behind this effect in our analysis. Emotional reactions in response to threat and episodic framing only occur among certain individuals (e.g., high in need for affect) and differ in nature between respondents (e.g., depending on levels of self-efficacy) meaning that the effect we demonstrate is not the same for everyone. Overall, the present study tackles a current and increasingly important topic with high social relevance, i.e. climate change and global warming. It is based on a large, representative sample and combines three different aspects, i.e. the role of threat framing, episodic vs. thematic framing and the role of visuals, previously looked at only in isolation. The interaction between these three research fields offers exciting analysis opportunities. Furthermore, we engage in an elaborate moderated mediation analysis in order to precisely define the conditions under which (i.e. moderators) as well as the underlying mechanisms how (i.e. mediators) these effects operate. Finally, we employ innovative measures how to assess emotional responses to news texts (e.g., IPANAT) and show their applicability, which are far from being established in current communication research.
HUMAN INTEREST FRAMED TELEVISION NEWS AND ITS EFFECT ON SUPPORT FOR GOVERNMENT INTERVENTION - MARK BOUKES, HAJO G. BOOMGAARDEN, MARJOLEIN MOORMAN AND CLAES H. DE VREESE The consequences of soft news and infotainment for the quality of democracy are heavily debated. Some scholars are positive (Bird, 2011; Van Zoonen, 2005; Zaller, 2003), whereas others mainly assume negative effects (Blumler, 1992; Thussu, 2007). The present experimental study fuels this debate with empirical evidence on how one particular aspect of news entertainization (Baum, 2003; Machin & Papatheoderou, 2002), that is human interest framing, affects people’s support for government intervention. Using a human interest frame, journalists explain a topic by portraying a person as a specific illustration of a broader issue (Semetko & Valkenburg, 2000). This is strongly related to the concept of episodic framing (Iyengar, 1991), which means describing topics with specific examples, whereas thematically framed news puts a topic in a broader context. It has repeatedly been shown that exposure to news with human examples leads to attributions of individual responsibility (Gross, 2008; Iyengar, 1991; Strange & Leung, 1999). Human interest framing 253
254
ETMAAL 2013
BOOK OF ABSTRACTS
makes it less likely to see the bigger picture and to perceive problems as society-wide phenomena (Costera Meijer, 2003; Iyengar, 1991; Rucinski, 1992). Therefore, human interest framing will make it less likely that people hold society responsible for solving the framed problem, which decreases people’s support for the government to intervene (Hannah & Cafferty, 2006). Furthermore, presenting individuals in unpleasant conditions as not being responsible for their problem, evokes more empathy than when these individuals are perceived as being responsible for their problems themselves (Betancourt, 1990; Schmidt & Weiner, 1988). When people see human interest frames, it is likely that responsibility is attributed to individuals; subsequently, this decreases empathy and decreases the perceived need for government intervention (Hannah & Cafferty, 2006; Schmidt & Weiner, 1988). Hypothesis 1: Exposure to a television news item framed in a human interest way leads to less public support for government intervention than exposure to the same news item framed in a substantial policy way. Political knowledge has been shown to be an important moderator of framing effects in previous studies (Chong & Druckman, 2007). After controlling for prior attitudes and exposure, political knowledge seems to facilitate framing effects (Druckman & Nelson, 2003), because it eases sense making of information presented in news items, enables effectively storing this in and efficiently retrieving it from memory (Krosnick & Brannon, 1993). These are crucial processes, necessary for frames to have an effect. Hypothesis 2: The negative effect of exposure to human interest frames on public support for government intervention compared to exposure to substantial policy frames is stronger for individuals with more political knowledge. An experiment was conducted in order to test the causal relationships as hypothesized above. Two characteristics of the experimental design differ from previous experimental studies on this topic: first, the fact that video is being used instead of print, and secondly, the inclusion of a mixed frames condition as suggested by various content analysis studies. Four videos were produced relying on a 1 x 3 between-subjects (frame: human interest versus mix of human interest and substantial policy versus substantial policy) factorial design with control group. Participants, 252 in total and relatively representative for the Dutch population, were randomly assigned across the four experimental conditions. The videos were short news broadcasts of three news items. One of these items was manipulated and dealt with a proposed government intervention in health care policy. The videos differed in who was interviewed by the journalist: a mother of a child affected by the situation (human interest condition), a politician of an opposition party (substantial policy condition), or both the mother and the politician (mixed condition). The interviewees used the same arguments to oppose the intervention; logically, the mother spoke much more in personal terms than the politician did. Contrary to studies on newspapers (e.g., Aarøe, 2011; Gross, 2008; Iyengar, 1991), the results of this study show that framing a television news item in a human interest way does not seem to affect the level of support for the proposed government intervention across-the-board. This reminds us of the 254
255
ETMAAL 2013
BOOK OF ABSTRACTS
importance of the medium under investigation. However, when the moderating influence of political knowledge is taken into account, human interest framing leads to lower levels of support for government intervention, but only for people with high political knowledge levels. They were less supportive after seeing the human interest framed item than after seeing the substantial policy framed item. The difference in support between conditions was smaller as people had less political knowledge, and almost zero for those with below average political knowledge.
TEACHING THE COMPUTER TO CODE FRAMES IN NEWS: A SUPERVISED MACHINE LEARNING APPROACH TO FRAME ANALYSIS - BJORN BURSCHER, DAAN ODIJK, RENS VLIEGENTHART, MAARTEN DE RIJKE AND CLAES DE VREESE In most framing studies a human-centred approach to the coding of news frames is employed. Frames are coded by means of indicator questions in manual content analysis. Humans can be welltrained to code frames in news, and their performance can be kept track of until accuracy and reliability reach satisfactory stages. However, human coding is a time-consuming and costly proces. Computers, on the other hand, are naturally suited for the processing of large volumes of documents and the repetitiveness of coding frames. Therefore, in this paper, we introduce a computer-aided method for automatic frame coding, which in future research can make frame analysis more efficient. We test this method by applying it to the coding of four common frames in Dutch news coverage: Conflict, economic consequences, human interest and morality. Our method relies on supervised machine learning, which is an approach from the field of Artificial Intelligence. When applied to text analysis, supervised machine learning aims at automatically classifying large amounts of documents based on statistical analysis of text attributes in a smaller set of human-annotated documents (the training set). Provided with sufficient training data, a mathematical function (called classifier) creates a statistical model to code unseen news stories. In our study we make use of a large corpus of manually coded Dutch newspaper articles in order to train two different types of classifiers (Naive Bayes and Support Vector Machine), and let them categorize unseen news stories according to the occurrence of the four news frames mentioned above. In doing so, text attributes from the individual news stories are analyzed into a set of exploratory variables (called features). These features are used to predict whether or not unseen news items contain a certain frame. We compare coding performance of the two classifier types when using different kinds of features (word presence, word counts, tf-idf and parsimonious language models). Supervised learning as a method for coding frames combines human ability to understand and interpret news stories and the computer's ability to scan large amounts of texts systematically and reliably. It improves on manual questionnaire-based frame coding, because it is less time-consuming, and, once a classifier is trained with a small set of human-annotated documents, no human coders are needed anymore. Therefore, it significantly decreases financial costs for frame coding. Furthermore, it enables scholars to analyze entire populations of texts one is interested in. Our data consists of front page news stories of the two Dutch newspapers De Volkskrant and NRC Handelsblad for the period between 1995 and 2011. For each year, a random sample of news stories has been manually coded for the presence of the four news frames. To measure the extent to which 255
256
ETMAAL 2013
BOOK OF ABSTRACTS
the five frames appear in stories we used a series of 16 questions to which human coders had to answer yes(1) or no(0). These questions function as indictors of the frames we are interested in. Parts of the coded news stories have been used to train the classifier, while others are used as a test set to compare the computer's codings with those of human coders, which we consider the gold standard. When comparing computer-based codings with human codings, we find high levels of accuracy for some frames, while for other frames our machine learning based approach is inferior to human coding. Generally, performance of our approach is highest for coding conflict and lowest for coding the morality frame. Furthermore, coding performance is highly dependent on which combinations of classifiers and features are employed. We discuss our results in terms of its implication for the study of framing.
FRAME DYNAMICS AND COMPLEXITY IN THE FINANCIAL CRISIS - DIRK OEGEMA, FRIEDERIKE SCHULTZ, JAN KLEINNIJENHUIS, WOUTER ATTEVELDT AND SABINE GEERS Crises such as the financial crisis (2007-now) are events that disrupt and change the social order and legitimacy of institutions fundamentally. They trigger feelings of uncertainty, framing processes and different interpretations in the public and news media, but also institutional attempts to communicatively re-establish social order and legitimacy (Patriotta, Gond & Schultz, 2011; Heath, 2004: 168). Although crises have real origins, they are constituted in the communicative interplay between public actors such as media, protesters, and organizations. Crises might be not only reduced, but even increased through attempts to react to the disruptive event. For example, as public’s dependence on media content is higher in crises, media reporting on public fear in the news could accelerate market panic in ways that compound economic disruption (Kleinnijenhuis et al., 2011 for the financial crisis). Also, a decrease in the amount and diversity of issues discussed (complexity) during the crisis might not only be a result of it, but even cause dramatic breakdowns in international stability (Suedfeld & Tetlock, 1977). Theory To explain such crises and the societal changes taking place in these situations, mass communication research focuses traditionally on the role of media and herein applied agendas or frames (e.g., Entman, 1993, 2007; Gamson & Modigliani, 1989; Benford & Snow, 2000; Goffman, 1974). Other disciplines such as management research focuses on the use of institutionalized frames in discourses, but furthermore discusses structural conditions like environmental and institutional complexity (e.g., Kostova & Zaheer, 1999). They acknowledge that the heterogeneity of organizations, reality constructions, and societal logics actors follow (Suddaby & Greenwood, 2005; Weber et al., 2008) influences societal change processes. So far, in research on crises and societal changes, the role of communicative complexity and dynamics, which can be defined as the complexity and dynamics of frames and agendas, are not investigated yet. Research started recently to observe dynamics of frames by studying them as networks of semantic associations (Baden, 2010; Schultz et al., 2011; Baden & Koch, 2011; Hellsten, Dawson, & Leydesdorff, 2010; Scheufele, 2006), for e.g. via semantic network analysis (Attefeldt et al, 2008). Management research, for example, started to theoretically discuss, how changes to an institution are conceived through frame shifting and frame blending, and how provisional reframing acquire a taken-for-granted quality (Cornelissen & Werner, forthcoming). 256
257
ETMAAL 2013
BOOK OF ABSTRACTS
We follow these attempts and explain the dynamics of the financial crisis by focusing on the following research question: How do the complexity and stability of agendas and frames change in the course of the crisis and how these relate to the financial markets? We study frame-complexity and –dynamics by analyzing the complexity of the semantic network of actors, issues and evaluations and their shifts over time. Our paper contributes to advance our understanding of framing, crisis communication and complexity, which is examined as a dynamic construct in the dimension of communication (communicative complexity). This paper studies the financial crisis in the USA and EU, as they are highly complex in nature, yet profoundly impact our politics, economy and society and hence receive abundant media coverage. Method To answer this question we conducted an automatic content analysis of newspaper articles that we collected from LexisNexis about the financial crisis between July 2007 and August 2012 from the US (New York Times, Washington Post and USA Today, N=50,000 articles) and Great Britain (Times, Mirror, and Guardian, N=60,000 articles). Results Agenda-complexity is operationalized as the entropy of the agenda computed per month; framecomplexity as the entropy of the distribution of co-occurrences (how often an object on the agenda occurs together with another object on the agenda in a text as graphically represented by the number of arrows in a semantic network). Results indicate that the complexity of the agenda and framing increases strongly when the dollar crisis changes into a crisis with the surfacing of the Greek budget problems, especially in British news. Agenda-stability (frame-stability) is operationalized by the correlation of the agenda (framing) this month with the agenda (framing) of last month (actually being the autocorrelation). A high correlation implies stability or little dynamism. The stability is relatively high in the US news. Agenda-stability sharply decreases in the British news with the onset of the Greek crisis and only recovers in 2012, although the framing stabilizes much sooner. Both complexity and stability are related to the stock market in the US and EU.
Session #7 Friday 8 February 11:45 – 12:45 hrs. Chair: Andreas Schuck Location: Heidelberg (M1-16) English session
WHICH POLITICIANS PASS THE MEDIA GATES AND WHY? A PRELIMINARY MODEL AND REVIEW OF PREVIOUS STUDIES - DEBBY VOS Within the trend of ‘mediatization’ and ‘personalization’ of politics, research has shown that the focus of news coverage has shifted from parties and organizations to individual politicians and leaders (e.g. Kepplinger, 2002; McAllister, 2007; Rahat & Sheafer, 2007). Appearing in news media is an important means for politicians to be successful in the contemporary candidate-centered politics. 257
258
ETMAAL 2013
BOOK OF ABSTRACTS
Citizens often rely only on news media to get informed about their representatives, which gives news media a crucial role in connecting voters to political actors. Media content helps the electorate to make knowledgeable choices in elections. Voters are likely to support only those candidates they know enough about and this knowledge is gained mostly from the media (Bartels, 1988). Therefore, if and how a politician appears in news media, can have important electoral consequences. Moreover, getting media coverage might be advantageous during the policy-making process. Legislators use the news media to gain public support and influence their counterparts in order to getting legislation passed (Kedrowski, 1996; Tresch, 2009. Although the coverage of individual politicians has increased, it remains a highly selective procedure of who becomes a news source. Since news media can’t cover all politicians equal, they follow certain norms or rules in deciding whom to select as a news source. In line with the increasing attention of news media for individual political actors, investigating media coverage of individual politicians is an expanding research field. Scholars have tried to identify those selection routines to explain why certain political actors receive more coverage than others do. But research about the determinants of the news coverage of politicians has been conducted with different concepts, operationalization and settings, which has resulted in inconsistent results. Surprisingly, these mixed outcomes did not instigate a real scholarly debate. Although this subfield within political communication is expanding and maturing, the basic question of who gets into the news and why lacks a conclusive answer. The goal of this paper is to provide an answer to this question: which politicians pass the media gates and why do they? First, I propose a preliminary model for explaining news coverage of politicians, thereby bringing some structure into the overwhelming group of possibly influential variables. I develop a typology consisting of three levels of determinants: the micro-level of politicians, the meso-level of news media and the macro-level of country characteristics. Within each level, several features influencing politicians news coverage are discussed. More importantly, I focus on interactions between these levels. Therefore, the main and basic level of the model consists of the micro-level of individual politicians, which is then combined with the meso- and the macro-level. Second, the proposed model is tested by reviewing previous studies concerning news coverage of individual politicians. Based on a systematic selection of 24 relevant studies, the variables in the typology are tested for their effect and the model as a whole for its validity. Results show that determinants on all three levels are important as well as interactions between determinants on the different levels. Concerning the micro-level, the political standing of a political actor proves to be the most important determinant for gaining news coverage. But also more personal characteristics of politicians have a positive effect on their coverage, such as their physical attractiveness. Politicians who lack these ‘qualities’ can enhance their news coverage by undertaking media activities such as sending press releases and maintaining good contacts with journalists. However, being politically very active -for example by proposing laws- has hardly a positive effect on the amount of coverage a politician receives. These effects on the micro-level are moderated by characteristics on the meso-level such as ownership of the medium and the type of event being covered. On the macro-level, the political and media system of a given country interact with microlevel variables of politicians, with journalists in different countries focusing on other characteristics of political actors when selecting news sources. To conclude, I argue for more precise and standardized 258
259
ETMAAL 2013
BOOK OF ABSTRACTS
measures of micro-variables, more focus on interaction patterns between levels and especially more comparative research between different countries.
WHEN POLITICS BECOMES NEWS: AN ANALYSIS OF PARLIAMENTARY QUESTIONS AND PRESS COVERAGE IN THREE WEST-EUROPEAN COUNTRIES - ROSA VAN SANTEN, PETER VAN AELST AND LUZIA HELFER When do journalists consider the words and actions of politicians newsworthy? This is a central question in the larger debate on the relationship between media and politics. There is little discussion that nowadays media coverage is crucial for a politician to get his/her message across and to reach out to voters. Furthermore, politicians can use media attention to obtain their political goals by influencing peers and eventually the policy process. For the news media, however, not all political actors and events are newsworthy. Not only because the “carrying capacity” of the media is too limited to report on all politicians do and say, but also because the media operate according to their own logic. They for instance focus on a few powerful political actors, or on highly conflictual events. This study aims to investigate the members of parliament (MP)-competition for media coverage by a detailed analysis of oral parliamentary questions (PQs) and press coverage thereof in three West European countries. We look for explanations of the newsworthiness of PQs on two levels of analysis. At the micro-level we study the most important characteristics of every parliamentary question. This includes the content of the question (voiced criticism, competence and causal attribution), the political actors involved (minister or state secretary addressed, coalition or opposition MP) and the media coverage that precedes the question. Next, a macro-analysis is added to see what institutional and system level characteristics of the Netherlands, France and Germany co-determine media coverage. We included factors such as whether or not PQs are asked and answered spontaneously, whether or not debate is allowed after the PQ, including additional questions from other MPs, and the total amount of questions asked in one question hour, might also determine the newsworthiness of PQs. We test hypotheses about these explanations through a content analysis of oral PQs and related newspaper coverage in three countries, in the most recent non-election year of each country. For each question we coded the actors involved, the issue, whether criticism was expressed, and whether a causal or competence attribution was made. Next, for each question the coverage in the national newspapers in the 5 days before and the two days after the question were analyzed. Articles that were about exactly the same topic as the PQ were coded. We collected 152 oral questions for the Netherlands (2011), 353 for France (2009) and 342 for Germany (2010). In the Netherlands a PQ most often was followed by media attention: 32% of the PQs received newspaper coverage after the question hour. In France 19% of the PQs received newspaper coverage, and in Germany only 16 questions led to media attention, which amounts to 5%. We conducted a Poisson regression in order to account for the right-skewed distribution of the dependent variable. Results show that the voicing of critique towards the addressed member of government, causal attribution and the volume of media attention before the question hour significantly affect the volume of media coverage of a PQ after the question hour.
259
260
ETMAAL 2013
BOOK OF ABSTRACTS
On the country-level, we find that in the Netherlands significantly more media attention is present after a PQ than in the other countries. Also, in the Netherlands an increase of media attention before positively affects the amount of media attention afterwards slightly more than it does in France or Germany. Other country-interaction effects were not significant, indicating that the effects of causal attribution and criticism do not significantly differ between the three countries. Results also show that the status of the actors involved does not seem to matter for the newsworthiness of a PQ. In all, we find the strongest results for the media-independent variable, which seems to indicate that newsworthiness is not so much determined by what a MP says, but by what (other) media outlets report on. We hope to shed some more light on this outcome and the country differences at the conference.
WHEN DO PROPOSED PENSION REFORMS LEAD TO A MEDIA WAVE? - ROEL DIJKMAN, WOUTER VAN ATTEVELDT, BARBARA VIS, LAURA HOLLINK AND PIM VAN DE PAVOORDT Media waves or media hypes are an important phenomenon in communication science but have received comparatively little attention in the literature. Building on the work of Wolfsfeld and Sheafer (2006) and Vasterman (2005), we conceptualize a media wave as a sudden rise in the amount of attention for a topic, leading to an intense discussion, gradually dying off. A key event triggers the rise of attention (Brosius and Eps 1995). In the case of political waves, the key event is generally initiated by a political actor (Wolfsfeld and Sheafer 2006). One drawback of existing studies on media waves is that they study only waves that actually happened. In other words, the key events that do not lead to media waves are not considered. This makes it difficult to determine when an event leads to a wave, and when not. To quantitatively study a sample of both waves and non-waves, one needs to be able to enumerate all key events of a specific kind, and then investigate whether these events lead to a wave or not. This study uses proposals by politicians to modify the Dutch public pension (AOW), either by changing the amount (e.g., taxing or freezing it) or by changing the pension age. These measures potentially affect the whole population at some point, so potentially generating a lot of discussion. For example, the 2006 proposal to fiscalise (i.e. tax) the AOW by social democrat's leader Bos sparked a wide debate in the media and was seen as a contributing factor to the subsequent poor election performance of the party. Interestingly, the very same measure was implemented after the elections without any real public debate. In this paper, we study the conditions under which proposals to reform the AOW lead to a media wave, and how large that media wave is. We draw our hypotheses from the literature on news values and agenda building. As stated by Vasterman, in order for an event to become a news wave, it has to be sufficiently newsworthy. News values research (Harcup and O'Neil, 2001) indicates that a numbers of aspects can make a story news worthy, including the medium's own agenda, surprise, relevance to the audience, and magnitude. A news wave triggered by a statement from a politician is a form of agenda building. Walgrave and Van Aelst (2006) found that the ability of politicians to set the media agenda is determined by a number of contingent factors, including medium type, issue type, political context (e.g., election time or not) and characteristics or the politician. Brandenburg
260
261
ETMAAL 2013
BOOK OF ABSTRACTS
(2002) found that in election times politicians are often able to set the media agenda. Hopmann et al. (2012) confirm this result, adding that it is especially the case for parties with coalition potential. From these findings, we hypothesize that the probability of triggering a news wave will be high if the proposal amounts to a large change, occurred during election time, and was made by a powerful politician. We further hypothesize that the wave will be greater if another powerful politicians reacts negatively to the proposal in the onset of the wave. These hypotheses are tested with a new dataset that combines political and media data. We first identify the political proposals done in parliament in the period 1945--2012, using an automatic content analysis. Next, we conduct a manual analysis to determine whether a statement indeed comprises a new reform proposal. For these new proposals, we code the extremity and direction of the proposal, and the power of the proposing politician. Then, we measure the amount of news about the proposals using automatic content analysis. Our preliminary results confirm that in election time political proposals generate more media attention. Explorative analysis also shows that this is especially the case in times of economic downturn, which is presumably linked to the magnitude of the proposal as harsher austerity or reform will be needed in difficult economic times. By studying this particular phenomenon, we gain more insights into when political proposals generate a large amount media attention. From a journalism studies perspective, this is important as a well functioning media system should give attention to those proposals that are relevant and important to the readership, not just to proposals made at the right moment. If we see contest over the news media as part of a more general struggle for political control (Wolfsfeld 1997), understanding when a reform proposal triggers a media wave gives insight into the position and role of the media in this political struggle.
ABSENT YET POPULAR? EXPLAINING NEWS VISIBILITY OF MEMBERS OF THE EUROPEAN PARLIAMENT - KATJANA GATTERMANN AND SOFIA VASILOPOULOU In the European Union (EU), the media represents the ‘intermediary’ that connects politicians to citizens by commenting on and informing about political activities and behaviour. Thus far, political communication research has comprehensively dealt with the news coverage of European election campaigns. However, few studies investigate the news coverage of European politicians once elected into office – even though that is when their decisions and behaviour directly influence EU politics and consequently affect the daily lives of EU citizens. This paper studies the visibility of Members of the European Parliament (MEPs) in national newspapers. It seeks to understand patterns of news visibility and identifies factors that are responsible for variation – both at the individual, party and domestic level. Thereby it concentrates on the media perspective taking both news values and external effects onto news coverage into account. The research sample comprises 316 MEPs from France, the Netherlands, Italy, the United Kingdom and Germany and covers a routine period of 25 months during the 7th parliamentary term, which
261
262
ETMAAL 2013
BOOK OF ABSTRACTS
began in July 2009. Both tabloids and broadsheets are selected for investigation. The paper measures news visibility with the ‘reading probability’ formula developed by Vliegenthart et al. (2005). The paper finds that at the individual level expertise and seniority in office have a positive effect on MEP news visibility. However, across the bench, less active MEPs and those known as ‘absentees’ receive more news coverage than their colleagues who more actively pursue their representative role inside the EP. On the other hand, those MEPs whose contributions in parliamentary debates via parliamentary questions and speeches exceed the legislative behaviour of the average MEP, which especially applies for Eurosceptic MEPs, are more visible in the news. One reason for that might be linked to the fact that these MEPs also have key roles in the domestic context. While party size is not a decisive factor, MEPs of national parties that form part of the national government are more visible in the news. In addition, at the domestic level, the distance to national general elections plays a crucial role for news visibility: The closer the polling date, the more visible MEPs are in the national press. Overall, although the paper is unable to test whether individual communication strategies of MEPs play a role, the results suggest that domestic politics matter more than European politics for MEP news coverage. This has implications both for our understanding of parliamentary behaviour at the EU level and for European communication research.
262
263
ETMAAL 2013
BOOK OF ABSTRACTS
Panels Meet the editors: de do’s en don’ts van het publicatieproces Thursday 7 February 14:05 – 15:05 hrs. Chairs: Jeroen de Keyser and Karin Fikkers Location: Forumzaal (M3-15) English session Jonge onderzoekers voelen meer dan ooit tevoren de druk om hun onderzoeksresultaten te publiceren in wetenschappelijke tijdschriften. Maar hoe kies je het juiste tijdschrift? Wat doe je als een review lang op zich laat wachten? En wat doe je wanneer een ingediend manuscript geweigerd wordt? Om die vragen te helpen beantwoorden verwelkomt het NeFCA Young Scholars Network alle jonge onderzoekers op deze speciale sessie over publiceren in academische tijdschriften. Enkele editors van belangrijke communicatiewetenschappelijke tijdschriften geven aan de hand van praktische voorbeelden een inkijk in de zaken waar een auteur op moet letten wanneer hij een manuscript indient. De sessie is interactief, dus we verwelkomen vragen vanuit het publiek. Het panel bestaat uit vijf editors van Nederlandstalige, Europese of internationale tijdschriften: - Connie de Boer (Tijdschrift voor Communicatiewetenschap) - Claes de Vreese (International Journal of Public Opinion Research) - Elly Konijn (Media Psychology) - Keith Roe (European Journal of Communication)
263
264
ETMAAL 2013
BOOK OF ABSTRACTS
Culture identity and media/ Culturele identiteiten en media: blijvende uitdagingen Thursday 7 February 15:15 – 16:15 hrs. Chair: Sofie Van Bauwel Location: Forumzaal (M3-15) Het hedendaags gefragmenteerde medialandschap stelt mediaonderzoekers voor een aantal pertinente uitdagingen. Niet langer is het mogelijk om een analytisch onderscheid te maken tussen de productie, het publiek, en de mediatekst. Mede door evoluties in technologie beperkt de rol van het publiek zich niet langer tot het consumeren van media. In tegendeel, interacties met en in media kennen een stijgend belang. Specifiek voor mediaonderzoek dat zich richt op het maatschappelijke belang van media voor culturele identiteiten, is het noodzakelijk om zich bewust te blijven van symbolische machtsdynamieken binnen zowel klassieke als nieuwe mediaomgevingen. Voor dit panel vertrekken we vanuit een cultureel theoretisch perspectief dat argumenteert dat media belangrijk zijn voor culturele identiteiten. Daarbij geldt niet enkel het belang van representatie, maar ook aanwezigheid en toegang tot media zijn primordiaal. Een cultureel theoretisch perspectief heeft hierbij voornamelijk aandacht voor een materiële en symbolische exclusie van identiteiten op basis van gender, seksualiteit, etniciteit en klasse. Dit panel heeft tot doel het blootleggen en het onderzoeken van deze mechanismen van maatschappelijke exclusie in het gefragmenteerd medialandschap. Dit doen we aan de hand van enkele case studies gebaseerd op empirisch onderzoek vanuit een Belgisch- Nederlandse context. Enerzijds gaat dit panel dieper in op deze vraagstukken in relatie tot klassieke media, zoals programmering binnen de Nederlandse publieke omroep en consumptie van nieuws in Vlaanderen. Anderzijds gaat dit panel dieper in op identiteiten in relatie tot nieuwe media, zoals gaming en sociale netwerksites. Daarbij wordt specifiek ingegaan op gender, seksuele, etnische, en klasse identiteiten. Vervolgens wordt ook de analytische bruikbaarheid van het concept “identiteit” zelf in vraag gesteld in relatie tot het alternatief “identificatie”.
264
265
ETMAAL 2013
BOOK OF ABSTRACTS
Celebrity and fan culture Thursday 7 February 16:45 – 17:45 hrs. Chair: Stijn Reijnders Location: Forumzaal (M3-15) De laatste jaren neemt het onderzoek naar zowel celebrities als fans in Vlaanderen en Nederland sterk toe. Dit groeiende Vlaams-Nederlandse onderzoeksveld biedt hiermee een antwoord op de dominante stroming in celebrity- en fanstudies die erg Angelsaksisch gericht is . Het doet dit door inzichten te bieden in deze maatschappelijk-culturele fenomenen vanuit een (vergelijkend globaal -) lokaal perspectief. Deze high density session verzamelt recent Vlaamse en Nederlandse onderzoek in celebrity- en fanstudies. De gemeenschappelijke noemer van deze bijdragen is dat ze zich begeven op de minder begane paden in celebrity- en fanstudies. Hier niet de focus op de rol van celebrities in de consumptiesamenleving, op de glamour van de grote internationale sterren, of op de brede groepen van mainstream fans of internationaal erkende ‘geeks’. Daarentegen belicht deze high density sessie tot dus ver minder verkend terrein, de zogenaamde ‘achterkant’ van de celebrity- en fancultuur, en dit op twee manieren. Enerzijds wil de sessie een blik werpen ‘achter de schermen’ door inzicht te bieden in het proces van celebrification in een lokale context (cf. bijdrage 1), in de manier waarop mainstream media de celebrity- en fancultuur helpen bouwen en continueren en tegelijkertijd presenteren als een afwijking van de norm in een complexe dialectiek van mainstream en marginale culturen (cf. bijdrage 2), en in de betekenis van ‘afgeleide’ processen van fandom zoals het verzamelen van informatie en objecten verbonden aan het object van de fandom (cf. bijdrage 3). Anderzijds wil de sessie aandacht schenken aan de ‘donkere kant’ van de celebrity- en fancultuur. Hierbij wordt aandacht besteed zowel aan de zogenaamde ‘infamous’ of ‘beruchte’ beroemdheden wiens claim to fame op negatieve eerder dan positieve handelingen en kenmerken zijn gebaseerd (cf. bijdrage 4), als ook aan de tegenpool van de fan, de ‘antifan’ die een intense maar negatieve relatie met de celebrities aangaat (cf. bijdrage 5). Via deze nadruk op de ‘achterkant’ van de fenomenen celebrity en fandom wil deze high density sessie nieuwe concepten, onderzoekslijnen en inzichten in het veld introduceren. De specifieke lokale Vlaamse en Nederlandse context biedt hierbij de noodzakelijke toetsing en herformulering van inzichten en begrippen gebaseerd op globale en sterk Angelsaksische geïnspireerde fenomenen. Bijdragen 1. Iedereen beroemd? Bekende Vlamingen en hun zelfdefinitie in tijden van democratisering van celebrity 2. ‘Superfans’: Het beeld van fans en hun idolen in mainstream media 3. Collecting Captain Kirk 265
266
ETMAAL 2013
BOOK OF ABSTRACTS
4. Guilty Houses: Dealing with Places of Crime in Belgium (Dutroux), the United Kingdom (Wests), and Austria (Fritzl) 5. With fans like these, who needs enemies? Een analyse van online antifans van celebrities Iedereen beroemd? Bekende Vlamingen en hun zelfdefinitie in tijden van democratisering van celebrity Olivier Driessen Deze paper vertrekt van de premisse dat celebrity gedemocratiseerd is—in die mate dat de verwachting vandaag is dat zowat iedereen beroemd kan worden—om te onderzoeken wat de impact hiervan is op de zelfdefinitie van celebrity’s en op de gepercipieerde (machts)verhoudingen binnen de celebritycultuur. Dit gebeurt via diepte-interviews bij Bekende Vlamingen uit diverse sectoren. Op die manier krijgen we inzicht in de subjectieve identiteitsposities van beroemdheden in een snel veranderende en hoogst competitieve omgeving. Verscheidene auteurs hebben al gewezen op de devaluatie van meritocratie in celebritycultuur. Ze veronderstellen een toenemende disarticulatie tussen roem enerzijds en aangeboren talenten of verwezenlijkingen anderzijds. Er zou met andere woorden een verschuiving opgetreden zijn van ‘verworven celebrity’ naar ‘toegewezen celebrity’ (cf. Rojek). Deze veronderstelde democratisering van celebrity gaat er daarbij van uit dat de toegang tot de hall of fame voor ‘gewone’ mensen vergemakkelijkt is. Vooral (nieuwe) mediatechnologieën en –platformen zijn daarbij cruciaal, met een vooraanstaande rol voor het internet en reality TV. Toch moeten we kritisch blijven bij deze positieve geluiden en erkennen dat deze do-it-yourself-(DIY-)celebrity’s en reality-sterren vaak vooral leeggezogen worden door de media- en celebrity-industrie via erg strikte contracten, zonder dat hen een kans op een carrière op langere termijn wordt aangeboden. Het zet ook de verhoudingen binnen de celebrity-industrie onder druk omdat veel van deze reality-sterren free labor aanbieden en op deze manier acteurs en televisiegezichten uit de markt prijzen. Een ander gevolg is dat de plotse toename en de grote zichtbaarheid van de DIY-celebrities en reality-sterren ook het aanschijn van de celebritycultuur hebben beïnvloed. De perceptie dat iedereen beroemd kan worden, dat daarvoor geen talent of prestatie meer noodzakelijk is, heeft ook de notie ‘celebrity’ aangetast en een hiërarchie gecreëerd tussen ‘echte’ sterren en zij die slechts bekendheid genieten omwille van hun bekendheid. In deze paper wordt dit ook empirisch onderzocht door middel van diepte-interviews bij 27 Bekende Vlamingen (BV’s) uit de sport, televisie, radio, muziek en film. Zien zij zichzelf ook als Bekende Vlaming en wat betekent dit label voor hen? Is de notie BV volgens hen inderdaad geëvolueerd en hoe? Hoe zouden ze zichzelf definiëren? Op basis van de interviewdata blijkt dat de meeste respondenten aanvaarden dat anderen hen als een BV omschrijven, terwijl velen onder hen voor zichzelf deze notie verwerpen. De reden is dat het label BV volgens hen een negatieve connotatie heeft en beperkt zou moeten worden voor die categorie van personen die bekend zijn omwille van hun bekendheid (cf. Boorstin). Hun zelfdefinitie als beroemdheid of BV blijkt gebaseerd te zijn op (een combinatie van) voornamelijk drie criteria die refereren aan het individu, het publiek en de media.
266
267
ETMAAL 2013
BOOK OF ABSTRACTS
Het individu als criterium verwijst naar het belang van prestaties, persoonlijke eigenschappen en talent. Vooral sporters, muzikanten en acteurs halen dit criterium aan: ze willen bekend zijn omwille van hun prestaties, niet enkel omwille van hun zichtbaarheid. In die zin is roem niet hun doel, maar een bijkomstigheid. Enkele respondenten oordelen dat je een BV bent niet omwille van persoonlijke eigenschappen of biografie, maar wanneer je herkend wordt door (een deel van) het publiek of wanneer je tot het ‘collectieve geheugen’ behoort. Gerelateerd is het criterium dat je een BV bent wanneer je door ‘de helft’ van de mensen herkend wordt op straat of wanneer ze je aanspreken, zo oordeelt althans een BV in dit onderzoek. Wat betreft media als criterium, valt het op dat een aantal BV’s argumenteren dat je een bekendheid bent van zodra de roddelbladen en televisiemagazines je tot onderwerp nemen. Anderen reserveren een cruciale rol voor de televisie, specifiek prime time televisie. Volgens hen ben je pas echt een BV wanneer je in prime time op de buis komt. Over het algemeen blijkt ook dat, conform de verwachtingen op basis van de literatuur, de noties BV en celebrity volgens hen een diepgaande verandering ondergaan hebben met de opkomst van reality TV. In België was dit volgens de respondenten met de uitzendingen van het eerste seizoen van Big Brother in 2000. Het label BV heeft sindsdien een negatieve connotatie, zo oordelen de meesten. In de conclusie van deze paper wordt ook stilgestaan bij mogelijke sociale wenselijkheid die de antwoorden van de respondenten kan beïnvloed hebben. Zo is de strikte scheiding tussen ‘wij’ die talent hebben en prestaties leveren en ‘zij’ die niets kunnen in zekere zin slechts relatief aangezien beide groepen (noodgedwongen) zichzelf en hun producten moeten promoten en beiden ook onderdeel zijn van dezelfde celebrity- en promotie-industrie. ‘Superfans’: Het beeld van fans en hun idolen in mainstream media Nathalie Claessens en Hilde van den Bulck In november 2011 bracht de Vlaamse commerciële televisiezender VT4 (SBS) de eerste uitzending van de reeks ‘Superfans’. Dit programma toont 23 mensen die fan zijn van een variëteit aan celebrities en waarvan het ‘fan zijn’ een belangrijke rol speelt in de dagelijkse bezigheden. In deze bijdrage wordt een analyse gemaakt van het beeld dat het programma ophangt van superfans en hun idolen en wordt dit geanalyseerd als het resultaat van een complexe dialectiek van mainstream en afwijkende culturen. Media spelen een centrale rol in de articulatie van een celebritycultuur en de (re)presentatie van celebrities. Media- en daardoor publieksaandacht is de zuurstof die de celebritystatus van (aspirerende) beroemdheden helpt garanderen. Tegelijkertijd zien we dat beroemdheden ook belangrijk zijn voor het succes van de media. Zo hebben televisiestations er belang bij dat hun presentatoren, omroepers en andere schermgezichten een zekere celebritystatus en daaraan gerelateerde fanbasis kunnen opbouwen die de kijkeraandelen kunnen helpen garanderen in een sterk competitief landschap.
267
268
ETMAAL 2013
BOOK OF ABSTRACTS
Mainstream media zoals commerciële televisiestations helpen dus de celebrity- en fancultuur op te bouwen en te continueren. Tegelijk suggereert het programmeren van een ‘ironische’ reality reeks over ‘superfans’ dat deze worden beschouwd als een ‘afwijking’ van een of andere norm en met voldoende potentiële ‘spektakelwaarde’ om een publiek te boeien. Deze bijdrage brengt een analyse van de manier waarop fandom wordt voorgesteld en hoe deze voorstelling een impliciete of expliciete evaluatie van fans tegenover een mainstream norm van ‘normale’ aandacht inhoudt. Hiertoe worden de fans die in het programma aan bod komen geanalyseerd, ten eerste, volgens de fan typologie van Abercrombie en Longhurst (1998) die een onderscheid maakt tussen fans, ‘cultists’ en ‘enthusiasts’, afhankelijk van het niveau van interesse, specialisatie, sociale organisatie en materiële productiviteit van de fans. Daarnaast worden ze geanalyseerd aan de hand van de functies die het fan zijn in hun leven vervult, zoals in de constructie, communicatie en beoordeling van de identiteit van de fan. Tot slot wordt het beeld van de fans ook geanalyseerd vanuit het perspectief van de ervaring van hun idolen: zien zij de fans en hun fangedrag als een last, een inbreuk op hun privacy of als een toegevoegde waarde aan hun beroemd zijn. De eerste analyses tonen en beeld van de fans als ‘cultists’. We zien welhaast obsessieve wezens die ‘opstaan en gaan slapen’ met ideeën en handelingen in relatie tot hun idool – door sommigen aangegeven als een verslaving. Voor een aantal van hen is hun levensgeluk afhankelijk van hun idool. De fans hebben bovendien allemaal het gevoel dat ze een zeer intense band met de celebrity hebben die varieert van ‘familiegevoel’, over celebrity-als-ouder en fan-als-kind, tot diepgaande verliefdheid. Deze band gaat gepaard met vaak hevige emoties, die soms een bijna religieuze toets krijgt waarbij de ontdekking van hun idool als een ‘openbaring’ wordt beschreven. Zoals in Abercrombie en Longhurst beschreven, produceren deze fans ook fanmateriaal. De betekenis van dit fandom zoals in het programma gepresenteerd wordt evenwel sterk bepaald door het format (reality) en de toon (‘uitlachtelevisie’) van de programma’s. De kenmerken van de fans worden voorgesteld als ‘freaky’ en ‘extreem’ en als sterk afwijkend van de ‘norm’ van de gemiddelde appreciatie. Dit wordt contrastief benadrukt door de focus in het programma op de moeilijkheden die fans hebben in hun sociaal leven als gevolg van hun fandom. Hier is het vooral de omgeving van de fans die aan het woord komt en die zich vaak ‘zorgen maakt’. De omgeving representeert hier de mainstream, de norm waartegen de fans worden afgezet. Collecting Captain Kirk Geen beschrijving Guilty Houses: Dealing with Places of Crime in Belgium (Dutroux), the United Kingdom (Wests), and Austria (Fritzl) Maloe Sniekers en Stijn Reijnders This paper focuses on the private houses of three infamous criminals from different regions in Europe: Fred and Rosemary West (United Kingdom), Marc Dutroux (Belgium) and Josef Fritzl (Austria). By adopting a comparative perspective, we aim to find general tendencies in the cultural process of dealing with the physical-material locations associated with violent death and suffering. Expanding on the work of Armando, who introduced the concept ‘guilty landscape’, the term ‘guilty houses’ is developed to refer to the abodes of these dark celebrities. Guilty houses can be described 268
269
ETMAAL 2013
BOOK OF ABSTRACTS
as houses that have been the silent witnesses of death and suffering that activate the memory of the cruelties that took place within them, retaining a sense of guilt. In total, 32 interviews have been carried out with local authorities, neighbouring residents, local entrepreneurs in the tourist industry, and tourists. All the houses involved were studied on-site. Results show that - notwithstanding small local differences - the houses all fulfil a similar, problematic role within the collective memory of these affairs. Locations associated with highly-mediated horrific crimes, like the houses of the Wests, Dutroux and Fritzl, can grow out to be local icons. These physical-material representations are difficult to deal with by local communities. For the majority, the houses serve as a symbolic reflection of the cruel events that took place within them. Therefore, a strong need to forget can be identified, often accompanied by the decision to demolish the houses by the municipalities. However, at the same time these locations attract tourists and commercial tendencies can be identified involving these sites, making it difficult to forget. With fans like these, who needs enemies? Een analyse van online antifans van celebrities Hilde van den Bulck and Nathalie Claessens Despite the media and academic’s dominant presentation of celebrities as loved and admired idols, doted on by their fans, famous people are not always the subject of admiration and love. In some extreme cases, audience members are not lovers but ‘haters’. This paper aims to provide a systematic analysis of this relatively neglected issue in fan studies: the phenomenon of antifans. With some notable exceptions (e.g. Gray 2003), little research has been done on antifans. This begs the questions: who are they and who or what constitute the objects of their dislike? What is it that turns people into antifans? What kind of emotional involvement – if any - does antifandom imply? Is antifandom merely about the enjoyment of dislike – for instance as emotional relief - or do deeper moral or societal expectations, values and norms come into play? Preliminary answers to these questions are suggested in Gray’s (2003) work on antifans which he does not just consider the mirror-image of fans but, on the contrary, finds to share several characteristics with fans. Elaborating on Gray’s work, this contribution, first, provides a theoretical discussion of characteristics of antifans as such and in relationship to fans and nonfans. Second, this is operationalized by linking antifans to the characteristics of so–called parasocial relationships that many audience members maintain with celebrities and that form an important basis for fandom. Third, a combined quantitative and qualitative analysis of antifans’ online reactions to media articles about celebrities provides systematic empirical insight into the particulars of antifans. Results show that, first, in contrast to fans, antifans feel good when something bad happens to the celebrity they dislike: either they are happy because celebrities are in pain or they want to cause that pain themselves (by using physical violence). Second, antifans are the opposite of fans in that they refrain from empathizing with the disliked celebrity and they make this unwillingness or inability to feel for the celebrity explicit in the comments. Third, antifans express their disliked very explicitly using curse words and expressing physical reactions (either wanting to hurt them or getting sick because of them), i.e. the mirror-image of fans. Results further show that not all antifans are the same. Based on the reasons behind the antifandom, three main types were identified: ‘antifans pur sang’, ‘fan antifans’ and ‘former fans’. The involvement of an antifandom pur sang is purely based on the dislike of a particular celebrity and 269
270
ETMAAL 2013
BOOK OF ABSTRACTS
what (s)he stands for. ‘Fan- antifan animosity’ is based on fans of one fan object becoming antifans of the opposing party that hurts the fan object. ‘Former fans’ emotions, finally, result from disappointment with the once adored fan object.
270
271
ETMAAL 2013
BOOK OF ABSTRACTS
Sloppy science Friday 8 February 11:45 – 12:45 hrs. Chair: Hilde van den Bulck Location: Forumzaal (M3-15) De nasleep van de Stapel affaire: Hoe “sloppy” is onze communicatiewetenschap eigenlijk? Format: Discussiepanel 15 min Korte presentaties (elk 5 min.) Verborgen flexibiliteit in Het replicatieprobleem dataverzameling en -analyse
Transparantie: Dataopslag en -deling
Anika Batenburg (VU Amsterdam)
Ivar Vermeulen (VU Amsterdam)
Camiel Beukeboom (VU Amsterdam)
45 min Publieke discussie a.d.h.v. de drie thema’s Discussieleider: Hilde van den Bulck (Universiteit Antwerpen) Discussanten: Jeroen Jansz (Erasmus Universiteit Rotterdam) Patrick de Pelsmacker (Universiteit Antwerpen) Rens Vliegenthart (Universiteit van Amsterdam
Achtergrond Onze taak als wetenschappers is om de werkelijkheid te ontdekken en te ontrafelen. Een aantal recente fraudegevallen binnen de sociale psychologie (bv. Diederik Stapel, Dirk Smeesters en Lawrence Sanna) heeft echter laten zien dat gedrevenheid om te scoren en zoveel mogelijk artikelen te publiceren dit wetenschappelijk doel soms overstijgt. In het eindrapport van de commissie Levelt over de Stapelfraude en in de daarop volgende persuitingen wordt niet alleen Stapel hard aangepakt; de gehele sociale psychologie maar ook belendende sociaalwetenschappelijke onderzoeksvelden worden beschuldigd van “sloppy science”. Recente literatuur lijkt aan te tonen dat onderzoekspraktijken die wetenschappelijke kennis ondergraven veelvoorkomend, en soms zelfs tamelijk geaccepteerd zijn (John, Loewenstein & Prelec, 2012). Dergelijke praktijken komen niet voort uit bewust bedrog, maar veeleer uit naïviteit, tijdsdruk, publicatiedruk en daarmee samenhangende zelfdienende interpretatie van wetenschappelijke bevindingen. Het gaat hierbij bijvoorbeeld over het opstellen/aanpassen van hypotheses na de dataverzameling (Kerr, 1998), het schoonpoetsen van data en het niet vermelden van nietsignificante resultaten, (Simmons, Nelson & Simonsohn, 2011), het niet uitvoeren van replicatiestudies (Francis, 2012), en onzorgvuldigheid in de opslag van data en rapportage van statistische analyses (Wicherts, Bakker & , Molenaar, 2011). In de nasleep van bovengenoemde affaires denken steeds meer wetenschappelijke tijdschriften, subsidieverstrekkers, wetenschappelijke afdelingen en zichzelf organiserende groepen onderzoekers na over mogelijke 271
272
ETMAAL 2013
BOOK OF ABSTRACTS
maatregelen om de wetenschapspraktijk te verbeteren. Dit geldt allang niet meer alleen voor de sociale psychologie, maar ook voor andere sociaalwetenschappelijke disciplines, zoals de communicatiewetenschap. Paneldiscussie In dit panel willen we de discussie onder communicatiewetenschappers over het gevaar van “sloppy science” binnen onze onderzoekspraktijk openen. Gelden bovengenoemde problemen ook voor de communicatiewetenschap? En zo ja, wat kunnen we er aan doen? Als inleiding zal in drie korte presentaties van circa 5 minuten elk een drietal centrale thema’s in de discussie over “sloppy science” worden besproken: (1) Verborgen flexibiliteit in dataverzameling en – analyse, (2) Het replicatieprobleem en (3) Transparantie: Dataopslag en –deling. Vervolgens zal de voorzitter de discussie met het aanwezige publiek openen, hierin bijgestaan door drie discussanten die hun visie op de onderzoekspraktijk vanuit verschillende inhoudelijke en methodologische invalshoeken kunnen belichten. Referenties Francis, G. (2012). The psychology of replication and replication in psychology. Perspectives on Psychological Science, 7, 585-594. John, L. K., Loewenstein, G., & Prelec, D. (2012). Measuring the prevalence of questionable research practices with incentives for truth telling. Psychological Science, 23, 524. DOI: 10.1177/0956797611430953 Kerr, N. L. (1998). HARKing: Hypothesizing After the Results are Known. Personality and Social Psychology Review, 2, 196. DOI: 10.1207/s15327957pspr0203_4 Simmons, J. P., Nelson, L. D., & Simonsohn, U. (2011). False-positive psychology: Undisclosed flexibility in data collection and analysis allows presenting anything as significant. Psychological Science, 22, 1359-1366. Wicherts, J. M., Bakker, M., & Molenaar, D. (2011). Willingness to share research data is related to the strength of the evidence and the quality of reporting of statistical results. PLoS ONE, 6(11), e26828. doi: 10.1371/journal.pone.0026828
272
