More than 1 independent variable involve
Data presented in both rows and columns
1 dependent variable present
Independent variable can be at scale, ordinal and nominal level
Dependent variable must be on interval and ratio level
Data should be normally distributed
Contingency table formed
Two way ANOVA:
The two-way ANOVA compares the mean differences between groups that have been split
on twoindependent variables (called factors). The primary purpose of a two-way ANOVA is to
understand if there is an interaction between the two independent variables on the dependent
variable.
Conditions:
Main Effect and Interaction Effect
The results from a Two Way ANOVA will calculate a main effect and an interaction effect. The
main effect is similar to a One Way ANOVA: each factor’s effect is considered separately. With
the interaction effect, all factors are considered at the same time. Interaction effects between
factors are easier to test if there is more than one observation in each cell.
Example:
Effect of exercise and age on the breathing capacity
Hypothesis:
H0 = exercise and age both have no effect on breathing capacity
H1 = exercise and age both have effect on breathing capacity
Level of significance =0.05
Conditions for rejection:
Reject H0 if sig value ≤ 0.05
Test applied:
Two way anova
Figure 1: Data entry
Figure 2: Application of two way anova test.
Figure 3: Application steps.
Figure 4:Application steps.
Results:
Levene's Test of Equality of Error Variancesa
Dependent Variable: breathingrate
F df1 df2 Sig.
1.102 8 18 .406
Tests of Between-Subjects Effects
Dependent Variable: breathingrate
Source Type III Sum of df Mean Square F Sig.
Squares
Corrected Model 9321.407a 8 1165.176 193.005 .000
Intercept 79272.926 1 79272.926 13131.098 .000
exerciselevel 1111.630 2 555.815 92.067 .000
Age 6843.185 2 3421.593 566.767 .000
exerciselevel * age 1366.593 4 341.648 56.592 .000
Error 108.667 18 6.037
Total 88703.000 27
Corrected Total 9430.074 26
1. R Squared = .988 (Adjusted R Squared = .983)
Objective;
The objective of the study was to check the hypothesis that is there any interaction between age
and exercise to affect the breathing rate.
Analysis;
Two -Way ANOVA.
Methodology;
First of all hypothesis was made and test was decided according to data. As the data fulfill the
conditions of two-way ANOVA so, we apply this test. Then normality of the data was checked
by using SPSS software. Keys provided already above. After which we applied the two-way
ANOVA through SPSS. Results in the form of tables were collected and data was presented in
the form of graph.
Results;
Lavaenes test value of 0.46 shows that variance between the groups were equal. Exercise has sig
value less than 0.05 that is 0.00, age has also sig value of 0.000. exercise*age has sig value of
0.000. this shows that exercise and age has significant effect on breathing rate.
Conclusion;
It was concluded from the results that the alternative hypothesis was accepted for effect on
breathing and there was significant difference between the means of breathing rate of 3 groups.
Significant difference or interaction represented by graph in which 3 lines cut each other that
shows that there is significant interaction between age and exercise to have strong impact on
breathing rate.
MNOVA (multiple analysis of variance)
Introduction
Multivariate analysis of variance (MANOVA) is simply an ANOVA with several dependent
variables. That is to say, ANOVA tests for the difference in means between two or more groups,
while MANOVA tests for the difference in two or more vectors of means.
Definition:
The one-way multivariate analysis of variance (one-way MANOVA) is used to determine
whether there are any differences between independent groups on more than one continuous
dependent variable. In this regard, it differs from a one-way ANOVA, which only measures one
dependent variable
Assumption:
Data is normally distributed
More than 1 independent variables present
MANOVA assumes that there are linear relationships among all pairs of dependent
variables
Dependent variables exhibit equal levels of variance across the range of predictor
variables.
Example of MANOVA
Perform the analysis of variance test on the following data and analyses the means using the LSD
test with a 0.05 level of significant.
Treatments
Blocks 1 2 3 4 5 6
1 1 3 6 4 3 2
2 1 4 4 8 5 1
3 3 6 7 8 4 3
4 2 3 2 3 2 1
Solution:
Hypothesis:
Ho: there is no significant interaction
H1: there is significant interaction
Results:
Tests of Normality
Kolmogorov-Smirnova Shapiro-Wilk
Treatments Statistic Df Sig. Statistic df Sig.
Score 1 .283 4 . .863 4 .272
2 .260 4 . .827 4 .161
3 .214 4 . .963 4 .798
4 .304 4 . .811 4 .123
5 .151 4 . .993 4 .972
6 .283 4 . .863 4 .272
Multivariate Testsc
Partial Eta
Effect Value F Hypothesis df Error df Sig. Squared
Treatments Pillai's Trace .503 1.210 10.000 36.000 .317 .252
Wilks' Lambda .497 1.424a 10.000 34.000 .212 .295
Hotelling's Trace 1.013 1.621 10.000 32.000 .145 .336
Roy's Largest Root 1.013 3.647b 5.000 18.000 .019 .503
Tests of Between-Subjects Effects
Dependent Type III Sum of Partial Eta
Source Variable Squares df Mean Square F Sig. Squared
Corrected Model Score 51.833a 5 10.367 3.588 .020 .499
Blocks .000b 5 .000 .000 1.000 .000
Intercept Score 308.167 1 308.167 106.673 .000 .856
Blocks 150.000 1 150.000 90.000 .000 .833
Treatments Score 51.833 5 10.367 3.588 .020 .499
Blocks .000 5 .000 .000 1.000 .000
Error Score 52.000 18 2.889
Blocks 30.000 18 1.667
Total Score 412.000 24
Blocks 180.000 24
Corrected Total Score 103.833 23
Blocks 30.000 23
a. R Squared = .499 (Adjusted R Squared = .360)
b. R Squared = .000 (Adjusted R Squared = -.278)
9
8
Means of treatments
7
6
5
4
3
2
1
0
1 2 3 4 5 6
Treatments
Fig: Comparison between means of six treatments.
Conclusion:
Computed value of F is 3.588 and p-value is 0.020 that is less than our decided α value that is
0.05 so, we reject the H0 and may conclude that there is difference among the six treatments and
we apply post hoc.
Report
Objective;
The objective of the study was to check the null hypothesis that there is no difference among the
means of treatments.
Conditions for rejection of null hypothesis:
Reject H0 if sig value is less than 0.05
Analysis;
Multivariate analysis of variance (MANOVA).
Results;
Computed value of F is 3.588 and p-value is 0.020 for treatments. Means were 1.75, 4.0, 4.75,
5.75, 3.50 and 1.75 for six treatments respectively.
Conclusion;
It was concluded from the results that the null hypothesis was rejected that there is no difference
among the six treatments because there was significant difference between the six treatments that
was shown by alphabets above the bar graph after application of post hoc test.
