Using ROIs, we have been trying to correlate measures of brain activity to
reaction time performance on a Stroop task. Our question is regarding
whether we should use copes or zstats to correlate with RT data. We've
conducted analyses using mean ROI values from both copes and zstat maps.
In this particular data set, we observed that more correlations were
significant using zstats (7 areas p < .05) than using copes (1 area p
< .05). We're trying to understand why these differences exist, and if
one method is to be preferred over the other. If so, why?