ABSTRACT
The current concern about the replicability of group-comparison science is totally legitimate. But it is only half of the problem. The other half is the plethora of replicable group studies that make claims about the psychology, the mind/brain, of individual subjects. Many studies justify claims about individual psychology with averaged data and individual exceptions. They should be rejected. Journal reviewers and editors should consider much more carefully than they do just what is the object of inquiry of the study under review. Does the study make claims about a group, like an alcohol-marketing study? Is it ambiguous, both about group behavior (economics) and individual psychology, like the Kahneman and Tversky experiments? Or is it about individual subjects (subject pairs) as in the implicit-evaluation study? In the last case, probably the proper course is to ask for more research: no scientific paper that draws conclusions about individual psychology based solely on averaged data should ever reach publication. Only when there are no exceptions and significance tests are no longer required can we be satisfied that there has been a real advance in understanding the psychology of individual human beings.
