TEORI ONLINE
SERI TUTORIAL SPSS by Hendry http://teorionline.wordpress.com/ email : openstatistik@
[email protected] yahoo.co.id
MEMULAI REPEATED MEASURE ANOVA FOR t
[email protected] ASUMSI REPEATED MEASURE Variable berskala interval or ratio (continuous). Dependent variable memiliki Normally distributed . Asumsi Sphericity Asumsi ini mirip Uji Homokedastis. Sample di ukur merupakan sampel yang sama minimal 2 kali.
Model Datanya seperti ini ya gan…ini contoh untuk data pendapatan X0
X1
X2
X3
2.85
4.86
2.65
2.85
5.08
7.17
5.20
5.16
3.75
5.57
3.16
3.56
4.37
6.26
3.91
4.25
3.36
5.26
3.24
3.25
4.23
5.99
3.91
3.98
2.37
3.72
1.81
1.71
3.93
5.65
3.55
3.65
6.96
9.04
6.70
7.03
6.35
8.38
6.24
6.37
5.31
7.19
5.03
5.19
3.19
5.04
2.84
3.04
5.93
8.05
5.91
6.04
4.30
6.12
3.86
4.11
2.95
4.58
2.71
2.58
3.72
5.58
3.41
3.57
4.52
6.35
4.40
4.34
3.55
5.42
2.88
3.41
X0 adalah data sebelum perlakuan X1 adala data setelah perlakuan (minggu pertama) X2 adalah data setelah perlakuan (minggu ke 2) X3 adalah data setelah perlakuan (minggu ke 3)
PROSES UJI ANALYZE -> general Linier Model repeated Measures
Ketik Method pada kotak : within Subject, dan masukkan 4 pada number of level. Klik ADD Maksudnya : kita akan menguji membandingkan hasil pengukuran berdasarkan 4 kali pengukuran Masukkan PDP (yang berarti pendapatan) pada Box Measure Name, lalu Klik ADD
Klik DEFINE
Masukkan data ke Box Kanan seperti contoh dengan mengklik Tombol
Click
. Lalu akan tampil sbb :
masukkan "METHOD" factor dari dari "Factors:" ke "Horizontal "Horizontal Axis:" Click
. Sehingga akan menjadi seperti ini
Click
.
Click Masukkan factor "METHOD" dari "Factor(s) and Fa ctor Interactions:" Interactions:" box ke "Display Means for:" box. Tandai "Compare main effects" effects" checkbox dan dan Pilih "Bonferroni" dari dari menu drop-down drop-down Tandai "Descriptive statistics" dan "Estimates "Estimates of effect size" checkboxes checkboxes di "Display" area.
Click
Click
.
HASIL
General Linear Model Within-Subjects Factors Measure: PDP
3
Dependent Variable Trans_A Trans_B Trans_C
4
Trans_D
METHOD 1 2
Descriptive Statistics Trans_A Trans_B Trans_C Trans_D
Mean 3.6631
Std. Deviation 1.41740
N
5.4141
1.65365
32
3.2391 3.4047
1.58866 1.65346
32 32
32
b Multivariate Tests
Effect METHOD Pillai's Trace
Value 1 .0 0 0
Wilks' Lambda
.000 Hotelling's Trace 1744 174472 72.4 .4 Roy's Largest Roo1744 174472 72.4 .4 a. Exact statistic b. Design: Intercept Within Subjects Design: METHOD
F Hypothesis df a 1686567 3 .0 0 0
Error df 29.000
Si g . .0 0 0
Partial Eta Squared 1.000
1686567 a 1686 16865 567a
3 .0 0 0 3 .0 0 0
29.000 29.000
.0 0 0 .0 0 0
1.000 1.000
1686 16865 567a
3 .0 0 0
29.000
.0 0 0
1.000
b Mauchly's Test of Sphericity
Measure: PDP a
Epsilon Approx. Within Subjects Effe Mauchly's W Chi-Square METHOD .0 0 1 221.728
df 5
Si g . .0 0 0
Greenhous e-Ge e-Geis iss ser Huyn Huynhh-Fe Feld ldtt Lowe Lowerr-bo boun und d .555 . 58 3 .333
Tests the null hypothesis that the error covariance matrix of the orthonormalized transformed dependent variable proportional to an identity matrix. a. May be used to adjust the degrees of freedom for the averaged tests of significance. Corrected tests are dis the Tests of Within-Subjects Effects table. b. Design: Intercept Within Subjects Design: METHOD
hasil Uji Sphericity ternyata ternyata data melanggar asumsi ini. (sig < 0.05) Tests of Within-Subjects Effects Measure: PDP Source METHOD
Type III Sum of Squares Sphericity Assumed 96.864 Greenhouse-Geisse 96.864
df
Mean Square F 3 32.288 870.663
Si g. . 000
Partial Eta Squared . 966
1.666
58.152
870.663
. 0 00
. 96 6
96.864
1.749
55.386
870.663
. 0 00
. 96 6
1.000 93
96.864 . 037
870.663
. 0 00
. 96 6
Error(METHO Sphericity Assumed
96.864 3 . 44 9
Greenhouse-Geisse
3 . 44 9
51.637
. 067
Huynh-Feldt
3 . 44 9
54.216
. 064
Lower-bound
3 . 44 9
31.000
. 111
Huynh-Feldt Lower-bound
Tests of Within-Subjects Contrasts Measure: PDP Source METHOD
Type III Sum of Squares 1 3 .9 2 7
1
Mean Square 1 3 .9 2 7
F 3 5 5 .9 1 6
Sig. .000
Partial Eta Squared .920
2 0 .1 0 6
1
2 0 .1 0 6
1083.115
.0 0 0
.972
6 2 .8 3 2
1
6 2 .8 3 2
1173.090
.0 0 0
.974
Quadratic
1 .2 1 3 .575
31 31
.0 3 9 .019
Cubic
1 .6 6 0
31
.0 5 4
METHOD Linear Quadratic Cubic
Error(METHOD)
Linear
df
Tests of Between-Subjects Effects Measure: PDP Transformed Variable: Average Source Intercept Error
Type III Sum of Squares 1977.183 306.593
df 1
Mean Square 1977.183
31
9.890
F 199.915
Si g . .000
Partial Eta Squared .8 6 6
Estimated Marginal Means METHOD Estimates Measure: PDP 95% Confidence Interval METHOD 1 2
M ean 3.663 5.414
Std. Error . 25 1 . 29 2
Lower Bound 3.152 4.818
Upper Bound 4 .1 7 4 6 .0 1 0
3
3.239
. 28 1
2.666
3 .8 1 2
4
3.405
. 29 2
2.809
4 .0 0 1
Pairwise Comparisons Measure: PDP
(I) METHOD 1
2
3
Std. Error . 060
Si g . .000
4
.424* .258*
. 04 8 . 06 0
.000 .001
. 288 . 089
. 560 . 428
1
1.751*
. 060
.000
1.581
1.921
3
2.175*
. 047
.000
2.044
2.306
4 1 2
2.009* -.424*
. 001 . 048
.000 .000
2.007 -.560
2.012 - . 288
-2.175*
. 047
.000
-2.306
-2.044
1
-.166* -.258*
. 046 . 060
.007 .001
-.296 -.428
- . 035 - . 089
2
-2.009*
. 001
.000
-2.012
-2.007
3
.166*
. 04 6
.007
. 035
. 296
(J) METHOD 2 3
4 4
95% Confidence Interval for a Difference Lowe Lowerr Bou Boun nd Upper pper Boun Bound d -1.921 -1.581
Mean Difference (I-J) -1.751*
Based on estimated marginal means *. The mean difference is significant at the .05 level. a. Adjustment for multiple comparisons: Bonferroni.
Ini adalah hasil perbandingan data antara A, B, C, dan D A dengan B berbeda signifikan A dengan C berbeda signifikan A dengan D berbeda signifikan dst
a
Profile Plots
Estimated Marginal Means of PDP
5.5
5.0
s n a e M l a4.5 n i g r a M d4.0 e t a m i t s E 3.5
3.0 1
2
3
METHOD
.
4
