Hello, Aside from a priori hypotheses that one may have about height and extent differences, I was wondering whether it is statistcally appropriate to use cluster-based statistics after doing a small volume correction? I thought that it had been mentioned in a message a while back that it was not appropriate, but I have not been able to find that message. Any insights (either for or against) would be greatly appreciated. Cheers, Barry ~~~~~~~~~~~~~~~~~~~~~~ Barry Giesbrecht, PhD Center for Cognitive Neuroscience Duke University Durham, NC 27708 USA [log in to unmask]