Dear Dan,
This could be due to e.g. 1) motion artefacts, with one group affected to a larger extent 2) inappropriate masks, 3) accidentally relying on CSF files, as the sig. clusters are located in regions where GM density/volume should be low (and likely not part of the analysis when relying on an e.g. explicit intensity threshold). It would be necessary to look at the individual segmented files though to get a better impression of the data.
Best
Helmut
|