One step further in mistreatment research: Assessing the scalability of the Explicit Discrimination Scale among Brazilian working-age adult respondents

Abstract

Though the Explicit Discrimination Scale (EDS) has been subjected to extensive psychometric evaluation in Brazil, the instrument has yet to be comprehensively assessed among working-age adult respondents in the country. This study aimed to fill this knowledge gap. Data from around 1200 diverse members of a cohort investigation were used to examine: (1) the positioning of respondents along the continuum of the EDS latent trait; (2) how well the corresponding items represent the EDS construct map; and (3) the extent to which the EDS items follow their expected levels of intensity. We assessed these properties with Loevinger's H, Guttman errors, and Item Response Theory parameters. Findings suggest that two abridged versions of the instrument—but especially the eight-item EDS—may adequately arrange respondents along the latent trait continuum. Analyses also revealed that scale items are reasonably spread over the construct map, with some discrepancy between the expected levels of intensity and their empirical positioning in the corresponding plot. The shortened versions of EDS have good psychometric properties among Brazilian working-age adult respondents. In addition to examining the invariance of the EDS across multiple groups, future psychometric evaluations should assess the external validity of the scale.