The "-glm_output option in randomise creates model estimates. The beta's are *_glm_pe_tstat1.nii.gz", where "pe" = "parameter estimates". We are usually interested in contrast parameter estimates ("cope"), so "_glm_cope_tstat1.nii.gz". The pe files will have one value (volume) per parameter in the model, whereas the cope files will have a single value corresponding to the contrast. For example, if you have group, age and sex, you will get files like below, where the "pe" files have three volumes each:
From FSL list: Say you have a contrast C = [1 -1 0]', and for a certain voxel, you have beta = [1 4 2]'. The COPE is simply C'*beta = [1 -1 0] * [1 4 2] = 1*1 -1*4 + 0*2 = -3. This is what goes in the numerator of the t statistic.
To view group differences in mricron:
...