Exploring unexplained variation

Abstract

In a 1974 commencement address, Richard Feynman described scientific integrity as a kind of utter honesty, a kind of leaning over backwards to tell the whole truth. We argue that investigators could tell more of the truth and increase the value of their papers by highlighting and discussing unexplained variation, a major source of which is individual differences. An argument that unexplained individual differences must have many sources is presented, and means of representing that variation are illustrated. We believe that such a change in reporting of research results is likely to advance the progress of scientific psychology, but perhaps the most compelling argument for what we propose is simply that telling the whole story as fully as possible is good scientific practice. The Appendix provides two examples of what we are urging, taken from recent psychological literature.