Hello,
My experimental design is: I have two groups of participants - group G1 (N=8) and group G2 (N=11). Total participants = 19.
Each group participated in 22 sessions (within-subject factor) in which their dependent variable Y was measured.
I wanted to do a two-factor mixed ANOVA with between-subject factor Groups (G1, G2) and within-subject factor Sessions (repeated measurement count = 22) as independent variable, and measurement Y as dependent variable.
In SPSS, Levene's test showed that the homogeneity of variance was satisfied for most of the sessions (p > 0.05), and failed for 3 sessions (p < 0.05).
Considering that the homogeneity of variance was satisfied for most of the sessions (19 out of 22 sessions), - Would my ANOVA be robust in spite of violation of homogeneity for 3 sessions?
Is there a method through which I can verify that I do not cause Type1 or Type2 error by running ANOVA? I tried running ANOVA by making the sample size equal between the groups (by excluding 3 participants in G2). There was no difference in p value, effect size range from ANOVA results that included all participants. Is it a valid justification to say that the ANOVA results are robust in spite of the violation of homogeneity?
Thanks in advance,
Sri