The -log10(p) is signed by the sign of the contrast, so to get back to the p value you must compute 10^-abs(sig).
On 7/26/16 10:42 AM, Caroline Beelen wrote:
Dear FS team,
I tried to run GLM-analysis using aparc stats files. At first I created these files for a few subjects of interest (e.g. aparc_rh_thickness_stats.txt) and checked whether they looked ok, which they did (values were pretty similar). Then I used the following command (for example):
Mri_glmfit
-- table aparc_rh_thickness_stats.txt
-- fsgd gender_dys.fsgd (see attached; example of a fsgd-file)
-- C Int.mtx
-- C Dys.mtx
-- C Gender.mtx
-- glmdir roi.gender_dyslect.rh.thickness.glmdir
The file was created and it contained a sig.table.dat folder. However, the values in this folder are all quite low. According to information found on your website they should correspond to (-)log10(p) values. I found mainly values between -1 and 1, but even of -0.03, which would correspond to a p-value of 1.5 (non-existing). Could you please explain to me what I did wrong?
Thank you very much!
Kind regards,
Caroline
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer