DAFTAR PUSTAKA Anderson MJ. Screening Ingredients Most Efficiently with Two-Level Design of Experiment (DOE). http://www.computer.org/publications/dlib. html [10 Januari 2006] Aunuddin 1989. Analisis Data. Bogor : PAU Ilmu Hayat Institut Pertanian Bogor. Bingham D, Sitter RR. 1999. Minimum aberration two-level fractional factorial split -plot design. Technometrics 41: 62-70. Bingham D, Sitter RR. 2001. Design issues in fractional factorial split-plot experiments. Journal of Technology. 33: 2-15. Birnbaum A. 1959. On the Analysis of Factorial Experiments Without Replication. Technometrics 1: 343 -59. Box GEP, Hunter JS. 1961. The fractional factorial design part I., II Technometrics 3: 311-48. Box GEP, William HG, Stuard HJ. 1978. Statistics for Experimenter. New York: John Wiley & Sons inc. Cochran WG, Cox GM. 1957. Experimental Designs. Ed ke-2. New York: John Wiley & Sons inc. Daniel C. 1959. Use of Half -Normal Plots Interpreting Factorial Two-Level Experiment. Technometrics 1: 311-41. Fries A, William HG. 1980. Minimum aberration 2 k-p . Technometrics 22: 601-08. Gomez KA, Arturo GA. 1995. Prosedur Statistika untuk Penelitian Pertanian. Ed ke-2. Sjamsuddin E, Baharsjah JS, penerjemah; Jakarta: UI Pr. Terjemahan dari: Statistical Procedures for Agricultural Research. Hines WW, Montgomery DC. 1996. Probability and Statistics in Engineering and Management Science. Ed ke-3. New York: John Wiley & Sons inc. Huang P, Dechang C, Joseph OV. 1998. Minimum aberration two-level split-plot designs. Technometrics 410: 314-26. Kulahci M, Ramirez JG, Tobias R. Split-plot Fractional Design: Is Minimum Aberration Enough?. Journal of Quality Technology 38 : 56-64. Loeppky JL, Sitter RR. 2002. Analyzing Unreplicated Blocked or Split-Plot Fractional Factorial Designs. Journal of Quality Technology 34 : 229-43. Montgomery DC. 2001. Design and Analysis of Experiments. Ed ke-5. New York: John Wiley & Sons, inc. Musa MS. 1999. Perancangan dan Analisis Percobaan. Bogor : Jurusan Statistika Institut Pertanian Bogor. Myers RH. 1990. Classical and Modern Regression with Application. Boston : PWS KENT Publishing Company. Nembhard HB, Navin A, Mehmet A, Seong K. 2006. Design Issue and Analysis of Experiments in Nanomanufacturing. Handbook of Industrial and Systems Engineering.
Lampiran 1. Alias interaksi dua faktor untuk tiga rancangan 2IV7-2
Interaksi 2 faktor AB AC AD AE AF AG BC BD BE BF BG CD CE CF CG DE DF DG EF EG FG
D1 CF + ACDG + BDFG BF + ABDG + ACDFG BCDF + BCG + FG BCF + ABCDEG + DEFG BC + ABCDFG + DG BCFG + ABCD + DF AF + DG + ABCDFG ACDF + CG + ABFG ACEF + CDEG + ABDEFG AC + CDFG + ADG ACFG + CD + ABDF ABDF + BG + ACFG ABEF + BDEG + ACDEFG AB + BDFG + ACDG ABFG + BD + ACDF ABCDEF + BCEG + AEFG ABCD + BCFG + AG ABCDFG + BC + AF ABCE + BCDEFG + ADEG ABCEFG + BCDE + ADEF ABCG + BCDF + AD
Alias interaksi dua faktor D2 CF + BDEG + ACDEFG BF + CDEG + ABDEFG BCDF + EG + ABCEFG BCEF + DG + ABCDFG BC + DEFG + ABCDG BCFG + DE + ABCDEF AF + ABCDEG + DEFG ACDF + ABEG + CEFG ACEF + ABDG + CDFG AC + ABDEFG + CDEG ACFG + ABE + CDEF ABDF + ACEG + BEFG ABEF + ACDG + BDFG AB + ACDEFG + BDEG ABFG + ACDE + BDEF ABCDEF + AG + BCFG ABCD + AEFG + BCEG ABCDFG + AE + BCEF ABCE + ADFG + BCDG ABCEFG + AD + BCDF ABCG + ADEF + BCDE
D3 CDF + DEG + ABCEFG BDF + BCDEG + AEFG BCF + BEG + ACDEFG BCDEF + BDG + ACFG BCD + BDEFG + ACEG BCDFG + BDE + ACEF ADF + ACDEG + BEFG ACF + AEG + BCDEFG ACDEF + ADG + BCFG ACD + ADEFG + BCEG ACDFG + ADE + BCEF ABF + ABCEG + DEFG ABDEF + ABCDG + FG ABD + ABCDEFG + EG ABDFG + ABCDE + EF ABCEF + ABG + CDEG ABC + ABEFG + CDEG ABCFG + ABE + CDEF ABCDE + ABDFG + CG ABCDEFG + ABD + CF ABCDG + ABDEF + CE
54
Lampiran 2. Struktur pembentukan rancangan FF 2 5 − 2 No
Generator & defining relation
1
D = AB ; E = AB I = ABD = ABE = DE
2
D = AB ; E = AC I = ABD = ACE = BCDE (Resolusi III)
3
(sesuai kriteria resolusi maksimum dan minimum aberration) D = AB ; E = BC I = ABD = BCE = ACDE (Resolusi III) Isomorphic dari no 2 (A à B ; B à A)
4
D = AB ; E = ABC I = ABD = ABCE = CDE (Resolusi III)
5
(sesuai kriteria resolusi maksimum dan minimum aberration) D = AC ; E = AB I = ACD = ABE = BCDE (Resolusi III) Isomorphic dari no 2 (C à B ; B à C)
6
D = AC ; E = AC I = ACD = ACE = DE
Alias
Pengaruh yg dianalisis
A = BD = BE A = BD = BE = ADE B = AD = AE B = AD = AE = BDE C C = ABCD = ABCE = CDE D = E = AB D = AB = ABDE = E AC AC = BCD = BCE = ACDE BC BC = ACD = ACE = BCDE CD = CE CD = ABC = ABCDE = CE AB terpaut dengan D&E; AD terpaut dengan B A = BD = CE = ABCDE A = BD = CE B = AD = ABCE = CDE B = AD C = ABCD = AE = BDE C = AE D = AB = ACDE = BCE D = AB E = ABDE = AC = BCD E = AC BC = ACD = ABE = DE BC = DE BE = AD E = ABC = CD BE = CD AB terpaut dengan D; AD terpaut dengan B A = BD = ABCE = CDE A = BD B = AD = CE = ABCDE B = AD = CE C = ABCD = BE = ADE C = BE D = AB = BCDE = ACE D = AB E = ABDE = BC = ACD E = BC AC = BCD = ABE = DE AC = DE AE = BDE = ABC = CD AE = CD AB terpaut dengan D; AD terpaut dengan B; BC terpaut dengan E A = BD = BCE = ACDE A = BD B = AD = ACE = BCDE B = AD C = ABCD = ABE = DE C = DE D = AB = ABCDE = CE D = AB = CE E = ABDE = ABC =CD E = CD AC = BCD = BE = ADE AC = BE AE = BDE = BC = ACD AE = BC AB terpaut dengan D; AD terpaut dengan B A = CD = BE = ABCDE A = CD = BE B = ABCD = AE = ABCD B = AE C = AD = ABCE = BDE C = AD D = AC = ABDE = BCE D = AC E = ACDE = AB = BCD E = AB BC = ABD = ACE = DE BC = DE BD = ABC = ADE = CE BD = CE AB terpaut dengan E; AD terpaut dengan C A = CD = CE = ADE A = CD = CE B = ABCD = ABCE = BDE B C = AD = AE = CDE C = AD = AE D = AC = ACDE = E D = AC = E AB = BCD = BCE = ABDE AB BC = ABD = ABE = BCDE BC BD = ABC = ABCDE = BE BD = BE AD terpaut dengan C
55
Lampiran 2. (lanjutan) No 7
Generator & defining relation D = AC ; E = BC I = ACD = BCE = ABDE (Resolusi III) Isomorphic dari no 2 (AàB ; BàC ; CàA)
8
D = AC ; E = ABC I = ACD = ABCE = BDE (Resolusi III) Isomorphic dari no 4 (B à C ; C à B)
9
D = BC ; E = AB I = BCD = ABE = ACDE (Resolusi III) Isomorphic dari no 2 (AàC ; CàB ; BàA)
10
D = BC ; E = AC I = BCD = ACE = ABDE (Resolusi III) Isomorphic dari no 2 (CàA ; AàC)
11
D = BC ; E = BC I = BCD = BCE = DE Isomorphic dari no 1 (C à A)
12
D = BC ; E = ABC I = BCD = ABCE = ADE (Resolusi III) Isomorphic dari no 4 (A à C ; C à A)
Alias
Pengaruh yg dianalisis
A = CD A = CD = ABCE = BDE B = CE B = ABCD = CE = ADE C = AD = BE C = AD = BE = ABCDE D = AC D = AC = BCDE = ABE E = BC E = ACDE = BC = ABD AB = DE AB = BCD = ACE = DE AE = BD AE = CDE = ABC = BD AD terpaut dengan C ; BC terpaut dengan E A = CD A = CD = BCE = ABDE B = DE B = ABCD = ACE = DE C = AD C = AD = ABE = BCE D = AC = BE D = AC = ABCDE = BE E = BD E = ABDE = ABC = BD AB = CE AB = BCD = CE = ADE AE = BC AE = CDE = BC = ABD AD terpaut dengan C A = ABCD = BE = CDE A = BE B = CD = AE = ABCDE B = CD = AE C = BD = ABCE = ADE C = BD D = BC = ABED = ACE D = BC E = BCDE = AB = ACD E = AB AC = ABD = BCE = DE AC = DE AD = ABC = BED = CE AD = CE AB terpaut dengan E; BC terpaut dengan D A = ABCD = CE = BDE A = CE B = CD = ABCE = ADE B = CD C = BD = AE = ABCDE C = BD = AE D = BC = ACED = ABE D = BC E = BCDE = AC = ABD E = AC AB = ACD = BCE = DE AB = DE AD = ABC = CDE = BE AD = BE BC terpaut dengan D A = ABCD = ABCE = ADE A B = CD = CE = BDE B = CD = CE C = BD = BE = CDE C = BD = BE D = BC = BCDE = E D = BC = E AB = ACD = ACE = ABDE AB AC = ABD = ABE = ACDE AC AD = ABC = ABCDE = AE AD = AE BC terpaut dengan D&E A = ABCD = BCE = DE A = DE B = CD = ACE = ABDE B = CD C = BD = ABE = ACDE C = BD D = BC = ABCDE = AE D = BC = AE E = BCDE = ABC = AD E = AD AB = ACD = CE = BDE AB = CE AC = ABD = BE = CDE AC = BE AD terpaut dengan E; BC terpaut dengan D
56
Lampiran 2. (lanjutan) No 13
Generator & defining relation D = ABC ; E = AB I = ABCD = ABE = CDE (Resolusi III) Isomorphic dari no 4 (D à E ; E à D)
14
D = ABC ; E = AC I = ABCD = ACE = BDE (Resolusi III)
15
Isomorphic dari no 4 (D à E ; E à D) (B à C ; C à B) D = ABC ; E = BC I = ABCD = BCE = ADE (Resolusi III)
16
Isomorphic dari no 4 (D à E ; E à D) (A à C ; C à A) D = ABC ; E = ABC I = ABCD = ABCE = DE
Alias
Pengaruh yg dianalisis
A = BE A = BCD = BE = ACDE B = AE B = ACD = AE = BCDE C = DE C = ABD = ABCE = DE D = CE D = ABC = ABDE = CE E = AB = CD E = ABCDE = AB = CD AC = BD AC = BD = BCE = ADE AD = BC AD = BC = BDE = ACE AB terpaut dengan E ; AD terpaut dengan BC A = CE A = BCD = CE = ABDE B = DE B = ACD = ABCE = DE C = AE C = ABD = AE = BCDE D = BE D = ABC = ACDE = BE E = AC = BD E = ABCDE = AC = BD AB = CD AB = CD = BCE = ADE AD = BC AD = BC = CDE = ABE AD terpaut dengan BC A = BCD = ABCE = DE A = DE B = ACD = CE = ABDE B = CE C = ABD = BE = ACDE C = BE D = ABC = BCDE = AE D = AE E = ABCDE = BC = AD E = BC = AD AB = CD = ACE = BDE AB = CD AC = BD = ABE = CDE AC = BD AD terpaut dengan BC & E A = BCD = BCE = ADE A B = ACD = ACE = BDE B C = ABD = ABE = CDE C D = ABC = ABCDE = E D =E AB = CD = CE = ABDE AB = CD = CE AC = BD = BE = ACDE AC = BD = BE AD = BC = BCDE = AE AD = BC = AE AD terpaut dengan BC
57
Lampiran 3. Penggunaan SAS 9.1 untuk pembentukan struktur rancangan FF Tahapan pembentukan struktur rancangan FF dengan SAS 9.1 adalah sebagai berikut : 1. Pilih menu SOLUTIONS à Analysis à Design of Experiments
2. untuk membuat rancangan FF yang baru, pilih menu FILE à Create New Design à Two-level…
58
Lampiran 3. (lanjutan) 3. Klik Define Variables. Klik Add> untuk menentukan banyaknya faktor yang akan dicobakan, untuk contoh ini dipilih 5 faktor yang digunakan. Kemudian klik OK untuk kembali ke kotak dialog sebelumnya.
4. Klik Select Design. Pilih Fractional factorial designs pada show designs of type. Tentukan fraksi percobaan dengan memilih type rancangan yang tersedia, dalam contoh ini pilih rancangan dengan fraksi ¼.
59
Lampiran 3. (lanjutan) 5. Klik Design Details... untuk mengetahui struktur rancangan yang terbentuk. Pada Design Information dapat diketahui informasi tentang resolusi maksimum yang bisa dicapai, didapat resolusi III sebagai resolusi maksimum
6. Pada Confounding Rules didapatkan generator yang terpilih sebagai pembentuk struktur rancangan terbaik. Dengan menekan panah ke bawah pada Principal : ++ dapat ditentukan seperempat bagian yang mana yang akan dicobakan, hal ini berkaitan dengan fold over.
60
Lampiran 3. (lanjutan) 7. Pada Alias Structure dap at diketahui susunan pengaruh faktor yang saling terpaut.
8. Klik tanda silang untuk menutup kotak Design Details dan kembali pada kotak dialog Two-Level design spesifications. Klik tanda silang untuk mengahiri pembentukan struktur rancangan FF.
61
Lampiran 4. Penggunaan ADX SAS 9.1 untuk Pengacakan Rancangan FF. Teknik pengacakan pada rancangan FF dapat dilakukan dengan klik edit response kemudian memilih menu Design à Randomized Design...
Hasil pengacakan akan didapatkan seperti pada kotak berikut:
62
Lampiran 5. Pembentukan struktur rancangan FFSP 2 ( 2 + 3) −(0 + 2 ) No
Defining contrast subgroups AP
1
Q = AP ; R = AP I = APQ = APR = QR
2
Q = AP ; R = BP I = APQ = BPR = ABQR (terbaik menurut kriteria resolusi maksimum dan minimum aberration)
3
Q = AP ; R = ABP I = APQ = ABPR = BQR (Resolusi III)
4
Q = BP ; R = AP I = BPQ = APR = ABQR (Resolusi III) Isomorphic dari no 2 (A à B ; B à A)
5
Q = BP ; R = BP I = BPQ = BPR = QR
6
Q = BP ; R = ABP I = BPQ = ABPR = AQR (Resolusi III) Isomorphic dari no 3 (A à B ; B à A)
Alias A = PQ = PR = AQR B = ABPQ = ABPR = BQR AB = BPQ = BPR = ABQR P = AQ = AR = PQR Q = AP = APQR = R BP = ABQ = ABR = BPQR BQ = ABP = ABPQR = BR Q terpaut dengan R A = PQ = ABPR = BQR B = ABPQ = PR = AQR AB = BPQ = APR = QR P = AQ = BR = ABPQR Q = AP = BPQR = ABR R = APQR = BP = ABQ AR = PQR = ABP = BQ Tidak ada pengaruh utama utama lain A = PQ = BPR = ABQR B = ABPQ = APR = QR AB = BPQ = PR = AQR P = AQ = ABR = BPQR Q = AP = ABPQR = BR R = APQR = ABP = BQ AR = PQR = BP = ABQ Tidak ada pengaruh utama utama lain A = ABPQ = PR = BQR B = PQ = ABPR = AQR AB = APQ = BPR = QR P = BQ = AR = ABPQR Q = BP = APQR = ABR R = BPQR = AP = ABQ AQ = ABP = PQR = BR Tidak ada pengaruh utama utama lain A = ABPQ = ABPR = AQ R B = PQ = PR = BQR AB = APQ = APR = ABQR P = BQ = BR = PQR Q = BP = BPQR = R AP = ABQ = ABR = APQR AQ = ABP = ABPQR = AR Q terpaut dengan R A = ABPQ = BPR = QR B = PQ = APR = AQR AB = APQ = PR = BQR P = BQ = ABR = APQR Q = BP = ABPQR = AR R = BPQR = ABP = AQ AP = ABQ = BR = PQR Tidak ada pengaruh utama utama lain
Pengaruh yg dianalisis A = PQ = PR B AB P = AQ = AR Q = AP = R BP BQ = BR A = PQ B = PR AB = QR P = AQ = BR Q = AP R = BP AR = BQ yang terpaut dengan pengaruh A = PQ B = QR AB = PR P = AQ Q = QP = BR R = BQ AR = BP yang terpaut dengan pengaruh A = PR B = PQ AB = QR P = BQ = AR Q = BP R = AP AQ = BR yang terpaut dengan pengaruh A B = PQ = PR AB P = BQ = BR Q = BP = R AP AQ = AR A = QR B = PQ AB = PR P = BQ Q = BP = AR R = AQ AP = BR yang terpaut dengan pengaruh
63
Lampiran 5. (Lanjutan) No 7
Defining contrast subgroups AP Q = ABP ; R = AP I = ABPQ = APR = BQR (Resolusi III) Isomorphic dari no 3 (Q à R ; R à Q)
8
Q = ABP ; R = BP I = ABPQ = BPR = AQR (Resolusi III)
9
Isomorphic dari no 3 (A à B ; B à A) (Q à R ; R à Q) Q = ABP ; R = ABP I = ABPQ = ABPR = QR
Alias
Pengaruh yg dianalisis
A = PR A = BPQ = PR = ABQR B = QR B = APQ = APR = QR AB = PQ AB = PQ = BPR = AQR P = AR P = ABQ = AR = BPQR Q = BR Q = ABP = APQR = BR R = AP = BQ R = ABPQR = AP = BQ AQ = BP AQ = BP = PQR = ABR Tidak ada pengaruh utama yang terpaut dengan pengaruh utama lain A = QR A = BPQ = ABPR = QR B = PR B = APQ = PR = ABQR AB = PQ AB = PQ = APR = BQR P = BR P = ABQ = BR = APQR Q = AR Q = ABP = BPQR = AR R = BP = AQ R = ABPQR = BP = AQ AP = BQ AP = BQ = ABR = PQR Tidak ada pengaruh utama yang terpaut dengan pengaruh utama lain A = BPQ = BPR = AQR A B = APQ = APR = BQR B AB = PQ = PR = ABQR AB = PQ = PR P = ABQ = ABR = PQR P Q = ABP = ABPQR = R Q=R AP = BQ = BR = APQR AP = BQ = BR AQ = BP = BPQR = AR AQ = BP = AR Q terpaut dengan R
64
Lampiran 6. Penggunaan SAS 9.1 untuk pembentukan struktur rancangan FFSP Tahapan pembentukan struktur rancangan FFSP dengan SAS 9.1 adalah sebagai berikut : 1. Pilih menu FILE à Create New Design à Split-plot…
2. Klik Define Variables. Pada Whole Plot Factor Klik Add> untuk menentukan banyaknya faktor petak utama yang akan dicobakan,untuk contoh ini dipilih 2.
65
Lampiran 6 (lanjutan) Pada Sub -plot Factor klik Add> untuk menentukan banyaknya faktor anak petak yang dicobakan, untuk contoh ini dipilih 3. kemudian klik OK untuk kembali pada kotak dialog sebelumnya.
3. Klik Select Design. Pilih type rancangan yang diinginkan, untuk contoh ini dipilih rancangan dengan 8 run.
66
Lampiran 6 (lanjutan) 4. Klik Design Details... untuk mengetahui struktur rancangan yang terbentuk. Pada Design Information dapat diketahui informasi tentang resolusi maksimum yang bisa dicapai, didapat resolusi III sebagai resolusi maksimum. Defining relation yang terbaik adalah APQ=BPR dengan WLP {2,1,0}
67
Lampiran 7. Penggunaan SAS 9.1 untuk pembentukan pengacakan struktur rancangan FFSP Pilih Edit Respon kemudian klik Design à Randomize Design...
Hasil pengacakan akan terlihat sebagai berikut :
68
Lampiran 8. Hasil analisis regresi dengan metode forward selection untuk percobaan pada contoh kasus rancangan FF •
Tahap 1 : Pengaruh faktor E masuk ke dalam model, R 2 model = 76.44% Analysis of Variance Source
Sum of Squares
Mean Square
F Value
Pr > F
45.41
<.0001
Type II SS
F Value
Pr > F
Model
1
7921.000 0
7921.0000
Error
14
2442.0000
174.4286
Corrected Total
15
10363.0000
Variable
Intercept E
•
DF
Parameter Estimate
Standard Error
54.2500
3.3018
47089 .0000
269.96
<.0001
-22.2500
3.3018
7921.0000
45.41
<.0001
Tahap 2 : Pengaruh faktor B masuk ke dalam model, R 2 model = 91.49% Analysis of Variance Source
DF
Sum of Squares
Mean Square
F Value
Pr > F
69.89
<.0001
Type II SS
F Value
Pr > F
Model
2
9481.2500
4740.6250
Error
13
881.7500
67.8269
Corrected Total
15
10363.0000
Variable
Parameter Estimate
Standard Error
54.2500
2.0589
47089.0000
694.25
<.0001
B
9.8750
2.0589
1560.2500
23.00
0.0003
E
-22.2500
2.0589
7921.000 0
116.78
<.0001
Intercept
No other variable met the 0.0500 significance level for entry into the model.
69
Lampiran 9. Data percobaan pada contoh kasus rancangan FFSP
A
B
P
Q
R
S
T
U
y
-1
-1
1
-1
-1
1
1
1
-1 1 -1 1 -1 1 -1 1 -1 1 -1 1 -1 1 -1 1 -1 1 -1 1 -1 1 -1 1 -1 1 -1 1 -1 1 -1 1
-1 -1 1 1 -1 -1 1 1 -1 -1 1 1 -1 -1 1 1 -1 -1 1 1 -1 -1 1 1 -1 -1 1 1 -1 -1 1 1
-1 -1 -1 -1 1 1 1 1 -1 -1 -1 -1 1 1 1 1 -1 -1 -1 -1 1 1 1 1 -1 -1 -1 -1 1 1 1 1
1 -1 -1 1 1 -1 -1 1 1 -1 -1 1 1 -1 -1 1 1 -1 -1 1 1 -1 -1 1 1 -1 -1 1 1 -1 -1 1
1 1 -1 -1 -1 -1 1 1 1 1 -1 -1 -1 -1 1 1 1 1 -1 -1 -1 -1 1 1 1 1 -1 -1 -1 1 1
1 -1 1 -1 -1 1 -1 1 1 -1 1 -1 -1 1 -1 1 1 -1 1 -1 -1 1 -1 1 1 -1 1 -1 -1 1 -1 1
1.00 0.50 37.46 32.26 36.54 33.34 4.00 2.50 1.00 34.74 1.20 76.86 2.10 76.34 1.10 37.66 1.90 2.50 31.06 37.06 31.34 40.94 3.20 5.20 6.10 37.54 6.00 82.06 7.10 84.34 2.10 50.46
70
Lampiran 10. Struktur alias untuk rancangan pada contoh kasus FFSP
Struktur Alias A = APQS = AQRT = APRU = APRST = APQTU = AQRSU B = BPQS = BQRT = BPRU = BPRST = BPQTU = BQRSU P = QS = PQRT = RU =PRST = QTU = PQRSU Q = PS = RT = PQRU = PQRST = PTU = RSU R = PQRS = QT = PU = PST = PQRTU = QSU S = PQ = QRST = PRSU = PRT = PQSTU = QRU T = PQST = QR = PRTU = PRS = PQT = QRSTU U = PQSU = QRTU = PR = PRSTU = PQT = QRS AB = ABPQS = ABQRT = ABPRU = ABPRST = ABPQTU = ABQRSU PT = QST = PQR = RUT = RS = QU = PQRSTU AP = AQS = APQRT = ARU = ARST = AQTU = APQRSU AQ = APS = ART = APQRU = APQRST = APTU = ARSU AR = APQRS = AQT = APU = APST = APQRTU = AQSU AS = APQ = AQRST = APRSU = APRT = APQSTU = AQRU AT = APQST = AQR = APRTU = APRS = APQU = AQRSTU AU = APQSU = AQRTU = APR = APRSTU = APQT = AQRS BP = BQS = BPQRT = BRU = BRST = BQTU = BPQRSU BQ = BPS = BRT = BPQRU = BPQRST = BPTU = BRSU BR = BPQRS = BQT = BPU = BPST = BPQRTU = BQSU BS = BPQ = BQRST = BPRSU = BPRT = BPQSTU = BQRU BT = BPQST = BQR = BPRTU = BPRS = BPQU = BQRSTU BU = BPQSU = BQRTU = BPR = BPRSTU = BPQT = BQRS ABP = ABQS = ABPQRT = ABRU = ABRST = ABQTU = ABPQRSU ABQ = ABPS = ABRT = ABPQRU = ABPQRST = ABPTU = ABRSU ABR = ABPQRS = ABQT = ABPU = ABPST = ABPQRTU = ABQSU ABS = ABPQ = ABQRST = ABPRSU = ABPRT = ABAQTU = ABQRU ABT = ABPQST = ABQR = ABPRTU = ABPRS = ABPQU = ABQRSTU ABU = ABPQSU = ABQRTU = ABPR = ABSTU = ABPQT = ABQRS APT = AQST = APQR = ARTU = ARS = AQU = APQRSTU BPT = BQST = BPQR = BRTU = BRS = BQU = BPQRSTU ABPT = ABQST = ABPQR = ABRTU = ABRS = ABQU = ABPQRSTU
Pengaruh Faktor 12.87 3.14 28.82 0.80 1.81 0.92 -26.53 1.59 2.44 -9.98 27.84 0.22 0.15 1.57 5.87 0.84 2.59 -0.13 0.74 0.77 0.17 1.29 -0.98 0.49 0.37 0.67 -0.33 0.79 -9.16 -0.83 1.59
71
Lampiran 11. Hasil analisis regresi dengan metode forward selection untuk percobaan pada contoh kasus rancangan FFSP •
Tahap 1 : Pengaruh faktor P masuk ke dalam model, R2 model = 30.37% Analysis of Variance Source
Sum of Squares
Mean Square
F Value
Pr > F
13.08
0.0011
Type II SS
F Value
Pr > F
Model
1
6644.1628
6644.1628
Error
30
15236.0000
507.8628
Corrected Total
31
21880.0000
Variable
•
DF
Parameter Estimate
Standard Error
Intercept
25.2344
3.9838
20377 .0000
40.12
<.0001
P
14.4094
3.9838
6644.1628
13.08
0.0011
Tahap 2 : Pengaruh faktor AP masuk ke dalam model, R2 model = 58.71% Analysis of Variance Source
DF
Sum of Squares
Mean Square
F Value
Pr > F
20.62
<.0001
Model
2
12846.0000
6423.1791
Error
29
9033.6877
311.5065
Corrected Total
31
21880.0000
Variable
Parameter Estimate
Standard Error
Type II SS
F Value
Pr > F
Intercept
25.2344
3.12003
20377 .0000
65.41
<.0001
P
14.4094
3.12003
6644.1628
21.33
<.0001
AP
13.9219
3.12003
6202.1953
19.91
0.0001
72
Lampiran 11 (Lanjutan). •
Tahap 3 : Pengaruh faktor T masuk ke dalam model, R 2 model = 84.45% Analysis of Variance Source
Sum of Squares
Mean Square
F Value
Pr > F
50.69
<.0001
Model
3
18478.0000
6159.2053
Error
28
3402.4298
121.5153
Corrected Total
31
21880.0000
Variable
Parameter Estimate
Standard Error
Type II SS
F Value
Pr > F
Intercept
25.2344
1.94868
20377 .0000
167.69
<.0001
P
14.4094
1.94868
6644.1628
54.68
<.0001
T
-13.2656
1.94868
5631.2578
46.34
<.0001
13.9219
1.94868
6202.1953
51.04
<.0001
AP
•
DF
Tahap 4 : Pengaruh faktor A masuk ke dalam model, R 2 model = 90.5% Analysis of Variance Source
DF
Sum of Squares
Mean Square
F Value
Pr > F
64.34
<.0001
Type II SS
F Value
Pr > F
Model
4
19802.0000
4950.613 4
Error
27
2077.5920
76.9478
Corrected Total
31
21880.0000
Variable
Parameter Estimate
Standard Error
25.2344
1.5507
20377 .0000
264.81
<.0001
A
6.4344
1.5507
1324.8378
17.22
0.0003
P
14.4094
1.5507
6644.1628
86.35
<.0001
T
-13.2656
1.5507
5631.2578
73.18
<.0001
13.9219
1.5507
6202.1953
80.60
<.0001
Intercept
AP
73
Lampiran 11 (Lanjutan). •
Tahap 5 : Pengaruh faktor PT masuk ke dalam model, R2 model = 94.15% Analysis of Variance Source
DF
Sum of Squares
Mean Square
F Value
Pr > F
5
20599.0000
4119.8913
83.65
<.0001
Error
26
1280.5892
49.2534
Corrected Total
31
21880.0000
Model
Variable
Parameter Estimate
Standard Error
Type II SS
F Value
Pr > F
25.2344
1.2406
20377 .0000
413.71
<.0001
A
6.4344
1.2406
1324.8378
26.90
<.0001
P
14.4094
1.2406
6644.1628
134.90
<.0001
T
-13.2656
1.2406
5631.2578
114.33
<.0001
PT
-4.9906
1.2406
797.0028
16.18
0.0004
AP
13.9219
1.2406
6202.1953
125.92
<.0001
Intercept
74
Lampiran 11 (Lanjutan). •
Tahap 6 : Pengaruh faktor AT masuk ke dalam model, R2 model = 95.41% Analysis of Variance Source
DF
Sum of Squares
Mean Square
F Value
Pr > F
86.54
<.0001
Type II SS
F Value
Pr > F
Model
6
20875.0000
3479.1657
Error
25
1005.0514
40.2021
Corrected Total
31
21880.0000
Variable
Parameter Estimate
Standard Error
25.2344
1.1208
20377 .0000
506.86
<.0001
A
6.4344
1.1208
1324.8378
32.95
<.0001
P
14.4094
1.1208
6644.1628
165.27
<.0001
T
-13.2656
1.1208
5631.2578
140.07
<.0001
PT
-4.9906
1.1208
797.0028
19.82
0.0002
AP
13.9219
1.1208
6202.1953
154.28
<.0001
AT
2.9344
1.1208
275.5378
6.85
0.0148
Intercept
No other variable met the 0.0500 significance level for entry into the model.
75
