Hi,
I recently had to re-run some old analyses having installed 5.0.8. The first-level COPEs using 5.0.8 differ from those I get when I run 5.0.6.
I'm on Debian and using FSL defaults - with basic motion regressors, fieldmaps, voxel-wise physiological regressors and confound regressors.
Would any of the implemented changes from 5.0.6 to 5.0.8 be expected to result in different COPE estimates?
Cheers,
Andreas
|