afni.nimh.nih.gov

Rock, actually I was wrong in both of my previous responses!

I meant to say that the stimulus duration does matter when creating the interaction regressor in step 5 (not 4) where the re-convolved time series is filtered through 0s and 1s.

Gang