Abstract:

This study evaluated the ability of the Explicit Discrimination Scale (EDS) to produce comparable estimates among respondents according to gender, color/race, and socioeconomic status. Analysis was based on data from two studies with students from Brazilian public universities. An abridged version of the EDS with eight items was evaluated by the alignment method. Findings indicated violation of invariance between color/race and gender groups. Reports of discriminatory experiences had better comparability between socioeconomic status strata. This study showed that EDS should be used with caution, especially to compare discrimination estimates between respondents of different colors/races and genders. The observed violation of invariance reinforces the need for additional research examining whether such a scenario persists in larger and more diverse samples from Brazil.

Keywords:
Psychometrics; Bias; Surveys and Questionnaires; Social Discrimination