Thank you Donald, I will add the subjects to a 2nd level flexible factorial design, and I think I understand you correctly that the results should then match the results from the t-test of 1st level contrast images (except for one-tail two-tail differences). I appreciate you pointing out that the group comparisons are actually invalid using this method - thankfully this should not effect our hypothesis which will focus on the interaction but it is very useful to know.
Kind regards,
Nick Fallon
|