Hi Y'all, I've recently found an error in the way that mri_surfcluster reports the area of a cluster when using a group surface such as fsaverage. The bottom line is that the area reported is about 25% too big. However, the cluster-wise p-values reported by mri_surfcluster based on simulations from mri_glmfit ARE NOT AFFECTED. Volume clustering is unaffected.
I have fixed this bug in the Martinos Center "dev" environment. Part of this fix is to make old versions of mri_glmfit simulation incompatible with new versions of mri_surfcluster (and vice versa) for surfaces (ie, mri_surfcluster will exit with an error or will crash -- either way, it will not complete), as this could generate erroneous p-values. We will be propagating the fix to our stable version after the holidays, and make new distributions available.
More info and updates can be found here:
https://surfer.nmr.mgh.harvard.edu/fswiki/GroupAverageSurface
As always, sorry for the bug.
doug
ps. On the brighter side, I discovered this bug while implementing gaussian random field (GRF) corrections for multiple comparisons, so soon you will be able to run GRF in seconds instead of running monte carlo simulations over a weekend.