SAS has something called the "cubic criterion" cutoff for finding the most appropriate number of clusters. Does R have anything that would replicate that? I've been searching the lists and can't seem to find anything that would point me in the right direction. Thank in advance, Philip Bermingham