Dear SPM users,
I am aware that this question is not really new however despite reviewing
the list I don’t see through, may you can give me some support:
Performing a simple VBM (group A vs B) with a small value correction, I
would expect:
1) I tell SPM the region I am interested in by an appropriate mask (e.g.
the cerebellum+brainstem+BG).
2) I get the regions of significant difference based on the p-values
corrected! for the respective volume.
Instead I get a table with svc.corrected p-values of voxels representing
some local maxima within the cluster defined by the whole brain analysis
(puncorr<0.001).
However the total extent of the "corrected cluster" (the cluster defined by
p-svc.corrected<0.05 (FDR)) is not shown on the glass brain (it remains
unchanged).
Is there a work around to get the hole cluster, defined by
p-svc.corrected<0.05 (FDR)? If not, that is the reason? And how the results
of the SVC are then interpreted?
Is it common sense to conclude, that if I found some significant voxels
within a cluster of the wohlebrain anlaysis (p<0.001) to be significant at
pFDR<0.05 after SVC, the whole cluster can be regarded as significant in
the sense of SVC ?
thanks in advance! - best regards,
Sebastian Thees
|