Abstract
This note of caution concerns correlational procedures for analyzing data regarding theoretical models postulating that the same relationships hold both for any given S and over a sample of Ss. Two methods of correlation (between and within Ss) are available, but they do not necessarily give the same results.