Sex and power: Why sex/gender neuroscience should motivate statistical reform

Abstract

Towards the end of the last century, statistical reporting in medical research underwent substantial reform, with null hypothesis significance testing replaced with an estimation approach. Interestingly, this reform may have been largely motivated by the social costs of error within medical research, rather than simply scientific error per se. This chapter briefly reviews the benefits of the estimation statistical approach as a means to producing reliable information about nature and then describes how the current statistical method of null hypothesis significance testing specifically contributes to scientific error in sex/gender neuroscience. The potential social harm that can arise from such errors in this area of research is then highlighted. It is suggested that sex/gender neuroscience may therefore provide a valuable model to motivate, on ethical grounds, statistical reform within the psychological sciences.