Dear Kyle,
p-values (or q-values) are attributes of local maxima or clusters of a
field, so you can't compute them for any voxel - you might want to read
this whole thread about topological inference:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=SPM;48172147.1005
However, by default, SPM only displays in the results table for each
cluster, the 3 main local maxima that are more than 8mm apart - you can
change this in spm_list.m l.201-202 so that more local maxima are
displayed. Clicking the 'current cluster' button will also give you a
more detailed list (32 local maxima, 4mm apart). Or just type the
following at the MATLAB prompt when looking at results:
>> spm_list('List',xSPM,hReg,Inf,0);
Best regards,
Guillaume.
On 13/09/10 19:44, Kyle Burger wrote:
> Hello all,
>
> I am trying to determine what the pFDR values are of voxels in a
> cluster other than the peak voxel (that is reported on the output).
> I believe additional significant voxels in the same cluster are
> reported in a lighter (grey) font, is that correct? But I am unclear
> if it is feasible to determine pFDR values of non-peak voxels.
>
> I greatly appreciate you assistance in advance.
> Thanks!
> Kyle
--
Guillaume Flandin, PhD
Wellcome Trust Centre for Neuroimaging
University College London
12 Queen Square
London WC1N 3BG
