Sensitivity and Response Bias in Detecting Gender Discrimination

Judgments about the presence or absence of discrimination are generally made in a contextof considerable uncertainty. Here, we apply Signal Detection Theory, a framework thatmodels decisions under uncertainty, to attributions to discrimination in order to differentiatebetween sensitivity (i.e., the ability to differentiate between situations with and without cuesto discrimination) and response bias (i.e., a relative tendency to make liberal rather thanconservative judgments about the presence of discrimination) in such judgments. To do so,we developed a new paradigm and used it to examine participants’ attributions to genderdiscrimination by manipulating the gender of victims and observers and the harm of theaction (Experiment 1), the direction of gender discrimination (Experiment 2), thediscrimination context (Experiment 3) and participants’ decision task (Experiment 4).Participants’ response bias was consistent with cultural stereotypes about genderdiscrimination and was affected by the actual base rate of discrimination in the experimentalcontext which they picked up explicitly only in the context of discrimination. The harmimplied by the action in question affected participants’ response bias such that they weremore liberal in cases in which victims were harmed, with small corresponding differences insensitivity. Our flexible and broadly applicable new paradigm opens up new possibilities forfurther research on attributions to discrimination and their analysis in terms of sensitivityand response bias.