Practical No: 11
Name of Experiment:
To perform LSD
Question:
Carry out the analysis of variance from the given LSD design.
B 15 C 12 A 18
C 15 A 12 B 25
A 19 B 18 C 25
Working Expression:
Hypothesis Setting:
Null hypothesis (𝐻0T): There is no significant difference between treatments.
Null hypothesis (𝐻0R): There is no significant difference between rows.
Null hypothesis (𝐻0C): There is no significant difference between columns.
Alternative hypothesis (𝐻1T): There is significant difference between treatments.
Alternative hypothesis (𝐻1R): There is significant difference between rows.
Alternative hypothesis (𝐻1C): There is significant difference between columns.
Level of Significance (α) = 0.05
Calculation Table:
PRE_1 RES_1
15.33 -.33
11.00 1.00
18.67 -.67
15.67 -.67
12.33 -.33
24.00 1.00
18.00 1.00
18.67 -.67
25.33 -.33
UNIANOVA Values BY Row Column Treatment
/METHOD=SSTYPE(3)
/INTERCEPT=INCLUDE
/SAVE=PRED RESID
/POSTHOC=Row Column Treatment(TUKEY)
/PLOT=PROFILE(Row*Column Row*Treatment Column*Treatment) TYPE=LINE ERRORBAR=NO
MEANREFERENCE=NO
YAXIS=AUTO
/PRINT DESCRIPTIVE LOF
/PLOT=RESIDUALS
/CRITERIA=ALPHA(.05)
/DESIGN=Row Column Treatment.
Univariate Analysis of Variance
Between-Subjects Factors
Value Label N
There are three 1 Row 1 3
rows(r1,r2,r3) 2 Row 2 3
3 Row 3 3
There are three 1 Column 1 3
columns(c1,c2,c3) 2 Column 2 3
3 Column 3 3
There are three 1 A 3
treatments(A,B,C) 2 B 3
3 C 3
Descriptive Statistics
Dependent Variable: Values
There are three There are three There are three Std.
rows(r1,r2,r3) columns(c1,c2,c3) treatments(A,B,C) Mean Deviation N
Row 1 Column 1 B 15.00 . 1
Total 15.00 . 1
Column 2 C 12.00 . 1
Total 12.00 . 1
Column 3 A 18.00 . 1
Total 18.00 . 1
Total A 18.00 . 1
B 15.00 . 1
C 12.00 . 1
Total 15.00 3.000 3
Row 2 Column 1 C 15.00 . 1
Total 15.00 . 1
Column 2 A 12.00 . 1
Total 12.00 . 1
Column 3 B 25.00 . 1
Total 25.00 . 1
Total A 12.00 . 1
B 25.00 . 1
C 15.00 . 1
Total 17.33 6.807 3
Row 3 Column 1 A 19.00 . 1
Total 19.00 . 1
Column 2 B 18.00 . 1
Total 18.00 . 1
Column 3 C 25.00 . 1
Total 25.00 . 1
Total A 19.00 . 1
B 18.00 . 1
C 25.00 . 1
Total 20.67 3.786 3
Total Column 1 A 19.00 . 1
B 15.00 . 1
C 15.00 . 1
Total 16.33 2.309 3
Column 2 A 12.00 . 1
B 18.00 . 1
C 12.00 . 1
Total 14.00 3.464 3
Column 3 A 18.00 . 1
B 25.00 . 1
C 25.00 . 1
Total 22.67 4.041 3
Total A 16.33 3.786 3
B 19.33 5.132 3
C 17.33 6.807 3
Total 17.67 4.848 9
Tests of Between-Subjects Effects
Dependent Variable: Values
Type III Sum of
Source Squares df Mean Square F Sig.
Corrected Model 183.333 a
6 30.556 13.095 .073
Intercept 2809.000 1 2809.000 1203.857 .001
Row 48.667 2 24.333 10.429 .087
Column 120.667 2 60.333 25.857 .037
Treatment 14.000 2 7.000 3.000 .250
Error 4.667 2 2.333
Total 2997.000 9
Corrected Total 188.000 8
a. R Squared = .975 (Adjusted R Squared = .901)
Lack of Fit Tests
Dependent Variable: Values
Source Sum of Squares df Mean Square F Sig.
Lack of Fit 4.667 2 2.333 . .
Pure Error .000 0 .
Post Hoc Tests
There are three rows(r1,r2,r3)
Multiple Comparisons
Dependent Variable: Values
Tukey HSD
(I) There are three (J) There are three Mean Difference
rows(r1,r2,r3) rows(r1,r2,r3) (I-J) Std. Error Sig.
Row 1 Row 2 -2.33 1.247 .342
Row 3 -5.67 1.247 .081
Row 2 Row 1 2.33 1.247 .342
Row 3 -3.33 1.247 .203
Row 3 Row 1 5.67 1.247 .081
Row 2 3.33 1.247 .203
Multiple Comparisons
Dependent Variable: Values
Tukey HSD
95% Confidence Interval
(I) There are three rows(r1,r2,r3) (J) There are three rows(r1,r2,r3) Lower Bound Upper Bound
Row 1 Row 2 -9.68 5.01
Row 3 -13.01 1.68
Row 2 Row 1 -5.01 9.68
Row 3 -10.68 4.01
Row 3 Row 1 -1.68 13.01
Row 2 -4.01 10.68
Based on observed means.
The error term is Mean Square(Error) = 2.333.
Homogeneous Subsets
Values
Tukey HSD a,b
There are three Subset
rows(r1,r2,r3) N 1
Row 1 3 15.00
Row 2 3 17.33
Row 3 3 20.67
Sig. .081
Means for groups in homogeneous subsets are
displayed.
Based on observed means.
The error term is Mean Square(Error) = 2.333.
a. Uses Harmonic Mean Sample Size = 3.000.
b. Alpha = .05.
There are three columns(c1,c2,c3)
Multiple Comparisons
Dependent Variable: Values
Tukey HSD
(I) There are three (J) There are three Mean Difference
columns(c1,c2,c3) columns(c1,c2,c3) (I-J) Std. Error Sig.
Column 1 Column 2 2.33 1.247 .342
Column 3 -6.33 1.247 .066
Column 2 Column 1 -2.33 1.247 .342
Column 3 -8.67 *
1.247 .036
Column 3 Column 1 6.33 1.247 .066
Column 2 8.67 *
1.247 .036
Multiple Comparisons
Dependent Variable: Values
Tukey HSD
(I) There are three (J) There are three 95% Confidence Interval
columns(c1,c2,c3) columns(c1,c2,c3) Lower Bound Upper Bound
Column 1 Column 2 -5.01 9.68
Column 3 -13.68 1.01
Column 2 Column 1 -9.68 5.01
Column 3 -16.01 -1.32
Column 3 Column 1 -1.01 13.68
Column 2 1.32 16.01
Based on observed means.
The error term is Mean Square(Error) = 2.333.
*. The mean difference is significant at the .05 level.
Homogeneous Subsets
Values
Tukey HSD a,b
There are three Subset
columns(c1,c2,c3) N 1 2
Column 2 3 14.00
Column 1 3 16.33 16.33
Column 3 3 22.67
Sig. .342 .066
Means for groups in homogeneous subsets are displayed.
Based on observed means.
The error term is Mean Square(Error) = 2.333.
a. Uses Harmonic Mean Sample Size = 3.000.
b. Alpha = .05.
There are three treatments(A,B,C)
Multiple Comparisons
Dependent Variable: Values
Tukey HSD
(I) There are three (J) There are three Mean Difference
treatments(A,B,C) treatments(A,B,C) (I-J) Std. Error Sig.
A B -3.00 1.247 .240
C -1.00 1.247 .739
B A 3.00 1.247 .240
C 2.00 1.247 .415
C A 1.00 1.247 .739
B -2.00 1.247 .415
Multiple Comparisons
Dependent Variable: Values
Tukey HSD
(I) There are three (J) There are three 95% Confidence Interval
treatments(A,B,C) treatments(A,B,C) Lower Bound Upper Bound
A B -10.35 4.35
C -8.35 6.35
B A -4.35 10.35
C -5.35 9.35
C A -6.35 8.35
B -9.35 5.35
Based on observed means.
The error term is Mean Square(Error) = 2.333.
Homogeneous Subsets
Values
Tukey HSD a,b
There are three Subset
treatments(A,B,C) N 1
A 3 16.33
C 3 17.33
B 3 19.33
Sig. .240
Means for groups in homogeneous subsets are
displayed.
Based on observed means.
The error term is Mean Square(Error) = 2.333.
a. Uses Harmonic Mean Sample Size = 3.000.
b. Alpha = .05.
Profile Plots
Decision:
For Treatments, p-value = 0.466 > α = 0.05.
So, we accept H0T and reject H1T.
For row, p-value = 0.882 > α = 0.05.
So, we accept H0R and reject H1R.
And, for column p-value = 0.758 > α = 0.05.
So, we accept H0C and reject H1C.
CONCLUSION:
Therefore, there is no significant difference between treatments , there is no significant difference between rows.
Likewise, there is no significant difference between columns.