we have aquired blocked design fmri data from a single patient pre- and post-treatment. To analyze treatment effects I compared the two sessions on the first level using spm5. Regarding the correction for multiple comparisons: would it be reasonable to use an FWE corrected cluster wise threshold of .05 (min cluster size = 80)? Or would it be preferable to use an FDR corrected voxel wise threshold?
Thanks in advance & best regards