---

œƒf[ƒ^“Η‚έ‚±‚έ
> mm <- read.table("Box4_R.tab")@@¦¨ƒf[ƒ^ƒtƒ@ƒCƒ‹FBox4_R.tabCBox4_R.data
> mm
REP N V DATA
1 1 N0 V1 4430
2 1 N0 V2 3944
3 1 N0 V3 3464
4 1 N0 V4 4126
5 1 N1 V1 5418
6 1 N1 V2 6502
7 1 N1 V3 4768
8 1 N1 V4 5192
9 1 N2 V1 6076
10 1 N2 V2 6008
11 1 N2 V3 6244
12 1 N2 V4 4546
13 1 N3 V1 6462
14 1 N3 V2 7139
15 1 N3 V3 5792
16 1 N3 V4 2774
17 1 N4 V1 7290
18 1 N4 V2 7682
19 1 N4 V3 7080
20 1 N4 V4 1414
21 1 N5 V1 8452
22 1 N5 V2 6228
23 1 N5 V3 5594
24 1 N5 V4 2248
25 2 N0 V1 4478
26 2 N0 V2 5314
27 2 N0 V3 2944
28 2 N0 V4 4482
29 2 N1 V1 5166
30 2 N1 V2 5858
31 2 N1 V3 6004
32 2 N1 V4 4604
33 2 N2 V1 6420
34 2 N2 V2 6127
35 2 N2 V3 5724
36 2 N2 V4 5744
37 2 N3 V1 7056
38 2 N3 V2 6982
39 2 N3 V3 5880
40 2 N3 V4 5036
41 2 N4 V1 7848
42 2 N4 V2 6594
43 2 N4 V3 6662
44 2 N4 V4 1960
45 2 N5 V1 8832
46 2 N5 V2 7387
47 2 N5 V3 7122
48 2 N5 V4 1380
49 3 N0 V1 3850
50 3 N0 V2 3660
51 3 N0 V3 3142
52 3 N0 V4 4836
53 3 N1 V1 6432
54 3 N1 V2 5586
55 3 N1 V3 5556
56 3 N1 V4 4652
57 3 N2 V1 6704
58 3 N2 V2 6642
59 3 N2 V3 6014
60 3 N2 V4 4146
61 3 N3 V1 6680
62 3 N3 V2 6564
63 3 N3 V3 6370
64 3 N3 V4 3638
65 3 N4 V1 7552
66 3 N4 V2 6576
67 3 N4 V3 6320
68 3 N4 V4 2766
69 3 N5 V1 8818
70 3 N5 V2 6006
71 3 N5 V3 5480
72 3 N5 V4 2014

> summary(mm)
REP N V DATA
Min. :1 N0:12 V1:18 Min. :1380
1st Qu.:1 N1:12 V2:18 1st Qu.:4481
Median :2 N2:12 V3:18 Median :5825
Mean :2 N3:12 V4:18 Mean :5479
3rd Qu.:3 N4:12 3rd Qu.:6581
Max. :3 N5:12 Max. :8832

œˆφŽqŽw’θ
> mm$REP <- factor(mm$REP)
> mm$N <- factor(mm$N)
> mm$V <- factor(mm$V)

œBartlettŒŸ’θ
> bartlett.test(mm$DATA~mm$REP)

Bartlett test for homogeneity of variances

data: mm$DATA by mm$REP
Bartlett's K-squared = 0.1282, df = 2, p-value = 0.938

> bartlett.test(mm$DATA~mm$N)

Bartlett test for homogeneity of variances

data: mm$DATA by mm$N
Bartlett's K-squared = 38.7511, df = 5, p-value = 2.665e-07

> mm$V <- factor(mm$V)
> bartlett.test(mm$DATA~mm$V)

Bartlett test for homogeneity of variances

data: mm$DATA by mm$V
Bartlett's K-squared = 2.1845, df = 3, p-value = 0.535

œ‚Q—vˆφ Split-plot ‚Μ•ͺŽU•ͺΝ
> fm <- aov(DATA~N*V+Error(REP/N), data=mm) ¦uNv‚Ν‚PŽŸ—vˆφ
¦uVv‚Ν‚QŽŸ—vˆφ
> summary(fm)

Error: REP
Df Sum Sq Mean Sq F value Pr(>F)
Residuals 2 1082577 541288

Error: REP:N
Df Sum Sq Mean Sq F value Pr(>F)
N 5 30429200 6085840 42.868 1.950e-06 ***
Residuals 10 1419679 141968
---
Signif. codes: 0 `***' 0.001 `**' 0.01 `*' 0.05 `.' 0.1 ` ' 1

Error: Within
Df Sum Sq Mean Sq F value Pr(>F)
V 3 89888101 29962700 85.711 < 2.2e-16 ***
N:V 15 69343487 4622899 13.224 2.105e-10 ***
Residuals 36 12584873 349580
---
Signif. codes: 0 `***' 0.001 `**' 0.01 `*' 0.05 `.' 0.1 ` ' 1

œ‘½d”δŠr
> pairwise.t.test(mm$DATA, mm$REP, p.adj="bonferroni")

Pairwise comparisons using t tests with pooled SD

data: mm$DATA and mm$REP

1 2
2 1 -
3 1 1

P value adjustment method: bonferroni
> pairwise.t.test(mm$DATA, mm$N, p.adj="bonferroni")

Pairwise comparisons using t tests with pooled SD

data: mm$DATA and mm$N

N0 N1 N2 N3 N4
N1 0.54 - - - -
N2 0.12 1.00 - - -
N3 0.12 1.00 1.00 - -
N4 0.15 1.00 1.00 1.00 -
N5 0.16 1.00 1.00 1.00 1.00

P value adjustment method: bonferroni
> pairwise.t.test(mm$DATA, mm$V, p.adj="bonferroni")

Pairwise comparisons using t tests with pooled SD

data: mm$DATA and mm$V

V1 V2 V3
V2 1.00000 - -
V3 0.15331 1.00000 -
V4 2.7e-08 1.1e-06 0.00021

P value adjustment method: bonferroni

---