Nasim and Judith (and everyone else),
I've found and fixed a bug that was causing the surfcluster results for the mc and perm simulations of absolute "abs" contrasts.
To implement the fix, get Linux 64bit versions of these files and copy them to $FREESURFER_HOME/bin (make backups of the originals:). ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/mri_surfcluster ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/mri_glmfit
You can then run something like:
mri_glmfit-sim --glmdir yourglmdir --no-sim mc-z.abs
The "--no-sim" option tells it NOT to do the full simulation again, just use the CSDs that are there to recompute the cluster results for each contrast, so it should be done in a matter of minutes.
If you've used a permutation, then run
mri_glmfit-sim --glmdir yourglmdir --no-sim perm.abs
Please let me know if this fixes the problem.
doug