ANOVA

AFNI Message Board
1

Dear AFNI experts,

We have two groups and would like to run seed-based analyses to see how the functional connectivity of a region of interest is correlated with a behavioural empathy score. We used 3dGroupInCorr and found the correlations in each group separately and than the contrast between groups, but we would like instead to run the analyses on all subjects and find the interaction group * empathy score.

How could we do an ANOVA like that? Is there a way that we could still use the output from 3dGroupInCorr for this analysis and what files should we use?

Thanks,

Marie

2

Marie,

First, follow examples 5, 9 or 10 in afni_proc.py to perform seed-based correlation analysis for each subject. Then, use 3dttest++ or 3dMVM for ANCOVA.