ABSTRACT
In psychology, null hypothesis signifi cance testing (NHST; Cumming & Fidler, Chapter 11, this volume) and meta-analysis (MA) (Cooper & Dent, Chapter 16, this volume) have occupied advocates of statistical reform for decades. Hundreds of psychology journal articles criticize the former and encourage more widespread use of the latter. In medicine, NHST has similarly been admonished and MA promoted. Misuse and misinterpretation of NHST have been widespread in both disciplines. The alternative statistical practices advocated by reformers have been the same in both disciplines, too-estimation (effect sizes and confi dence intervals [CIs]) and, increasingly, MA.
