Learning to detect sexism: An evaluation of the effects of a brief video-based intervention using ROC analysis

Empirical evidence for the effectiveness of interventions teaching lay people how to recognize sexism is scarce. The purpose of the present study was, thus, twofold: The first aim was to evaluate a brief intervention using a lecture-like educational video on how to recognize subtle sexism. The second aim was to demonstrate the usefulness of signal detection theory (SDT) for evaluating the participants' ability to discriminate between subtle sexist and non-sexist statements. Participants (N = 73) were randomly assigned to a subtle sexism treatment group (SSG), an overt sexism treatment group (OSG), or a control group (CG). After the intervention phase, the participants were asked to rate statements in vignettes with respect to how sexist they perceived them to be. The participants in the SSG were significantly better in correctly identifying subtle sexist content than the participants in the OSG and CG. However, they were not more accurate overall. This was because they claimed sexism more often, irrespective of whether it was present or not. We conclude that while our intervention increased participants' sensitivity in detecting sexist content, it did so at the cost of specificity. Our results make clear that practitioners teaching people how to recognize sexism should control intervention outcomes for unintended effects of biased decision criteria, given that erroneous allegations of sexism could have grave consequences. To this effect, the value of SDT, which allows for fine-grained and, consequently, more accurate insight than standard approaches to the analysis of intervention effects, was demonstrated.