Dear everyone,
I have come across an fMRI paper presenting results of an FSL analysis. Unfortunately I'm not that familiar with FSL (usually work with SPM, shame on me ;-) ), so I would greatly appreciate if someone could help with a quite trivial question: According to the methods section, first-level (individual subject) models were thresholded at a certain Z and a certain cluster threshold. For group analysis, also certain thresholds are reported.
Now I'm a little curious: As far as I'm familiar with FSL GLMs the chosen threshold for first-level models does not affect group results, it's just useful when checking data (blobs present yes/no) and/or if one really wants to present individual results (which is not the case in that paper). However, as the first-level thresholds are reported, does this imply that the group model was conducted on a subset of voxels = the intersection of the individual "significance maps" then? Or why would one want to report those otherwise? Thank you very much!
Best regards
Helmut
|