Reporting Effect Sizes: the Roles of Editors, Textbook Authors, and Publication Manuals

In writing this article, I wear four relevant hats. I have served as a journal editor; I am the author of two undergraduate textbooks (Hyde, 1991; Hyde & DeLamater, 2000); I am chair of the Publications and Communications Board of the American Psychological Association, which supervises the revision of the Publication Manual; and I am a meta-analyst (e.g., Hyde, Fennema, & Lamon, 1990; Jaffee & Hyde, 2000; Kling, Hyde, Showers, & Buswell, 1999). I should begin by stating my personal position on the surprisingly controversial issue of effect size reporting. I believe that researchers should report the results of appropriate significance tests, such as F, and should report the effect sizes associated with each significance test (Wilkinson & APA Task Force on Statistical Inference, 1999). Each of these numbers carries somewhat different information, and readers have a right to all that information. Readers have a right to know, for example, how large an effect is, and researchers need to know the size of the effect if they are planning a follow-up study to decide on an appropriate sample size. I believe that such reporting is a minimum scientific standard. From this perspective, then, a discussion of who should lead the movement is not entirely sensible. It is simply a question of “doing the right thing,” and all constituencies should collaborate in maintaining a minimum scientific standard. From this perspective, too, the leading should come from graduate departments of psychology and other disciplines as they train the next generation of scientists in data-analytic techniques.