Measurement considerations in the link between racial/ethnic discrimination and adolescent well-being: A meta-analysis

Abstract

Racial/ethnic discrimination is a commonplace experience for many adolescents of color, and an increasing number of studies over the past 25 years have sought to document discrimination and its consequences at this stage of the life course. The evidence is clear and convincing that racial/ethnic discrimination is harmful for adolescents’ socioemotional and behavioral well-being as well as their academic success. Discrimination measurement, however, poses a critical source of potential variation in the observed effect sizes capturing the associations between racial/ethnic discrimination and adolescents’ well-being. This meta-analysis integrated 1,804 effect sizes on 156,030 unique ethnically- and geographically-diverse adolescents (Mage = 14.44, SD = 2.27) from 379 studies that used 79 unique instruments to assess racial/ethnic discrimination. The meta-analysis focused on a host of measurement-related moderators, including the number of items, response scale and response dimensions, reliability, retrospective reference period, perpetrators, and initial target populations. Larger effect sizes were observed for instruments with more items and with non-dichotomously rated items. Perpetrator and retrospective reference period also emerged as key moderators, while measure reliability, response dimensions, and initial measurement development characteristics were not significant moderators. Findings provide key insights for the development of more precise, effective instruments to assess perceived racial/ethnic discrimination in adolescence.