Hi,
I did a conjunction analysis of two contrasts, one set at p < 0.001 and the other at p < 0.05. When I run post-hoc t-tests on the ROI that I found with the conjunction I find p values that are less (more significant) than the thresholds (p< 0.001 and p< 0.05) used to identify the effects in the first place. (both the afni contrasts and the t-test I am using are 2 tailed).
Isn't that weird or am I misunderstanding something?
Thanks
Elisa