Dear colleagues,
If we have 2 groups: experimental and control, then we took samples in 5 phases. For within groups comparisons, we have to use repeated measure ANOVA. You know this test combines control and experiment groups data together in each phase, then it compares 5 phases. Also, it does not show between subjects comparisons. I have 2 questions:
1. In each group, how can we separately do within subjects comparison?
2. How can we compare control and experimental groups in each phase (between group comparison)?
Are there any reference for such a test?
Best Regards
Farzad
I assume you are talking about continuous variable as your outcome. If the outcome is binary, then story would be different.
You can use linear mixed model repeated measures analysis. Basically you will have phase., group, and phase by group interaction terms in your model. You can use linear contrast to test group effect at each phase, overall, and phase effect within each group. Comparing to two-way repeated ANOVA, you have option to specify which covariance structure you could use to incorporate the within subject correlation in the model.
Check this link:
http://www.theanalysisfactor.com/advantages-of-repeated-measures-anova-as-a-mixed-model/
ANOVA is the test to use. Two groups are involved. The total sum of squares is partitioned into (between) group sum of squares and error (I. e. within group) sum of squares. The between group comparison is made on the basis of f=gms/ems where gms is the between group mean squares and ems is the within group mean square. The ordered pair of degrees of freedom is (1, n1+n2-2) where n1 and n2 are the group sample sizes. The within group comparison is done by testing ems = 0 against ems positive.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling
- Analyses of Variance
- ANOVA