Dear Freesurfer experts,

My goal is to compare the cortical thickness between three groups. For that purpose, I performed whole brain analysis with a F-contrast to evaluate a group effect. Since the F-test was significant, I performed pairwise comparisons to evaluate the direction of these differences. However, I noticed that some brain areas that are significant in one pairwise comparison (that are consistent with our hypothesis) were not significant in the F-test. I am a bit lost about how to proceed in this situation, can I report significant results in the pairwise comparison that were not previously significant in the F-contrast or do I need to perform the pairwise comparisons by applying a mask to the brain areas where there were significant results in the F-test?

Apologies for asking this simple question, I am a beginner in statistical analysis with MRI data.

Thank you very much for your help.

Best regards,