Construct Validation of a Shortened Version of the Everyday Discrimination Scale

Aprile Benner , University of Texas at Austin
Shanting Chen
Celeste Fernandez

Given the population health implications for racial/ethnic discrimination, it is imperative that population-based studies integrate measures of racial/ethnic discrimination; however, including even relatively short measures (e.g., Everyday Discrimination Scale [EDS], 9 items) may not always be feasible due to time and space constraints. Using data from NSAL, NLAAS, and Add Health, we examined the construct validity of the 9-item versus a 2-item EDS scale (NSAL/NLAAS) and a single-item measure combining the first two EDS items (Add Health). The 9-item scale had item-total correlations ranging from .55-.74, and the correlation for the 2-item measure was .69. The correlation between the 9-item and the 2-item scale was .74, and these correlations were similar for men/women and across racial/ethnic groups. The three scales had similar relations to known correlates spanning psychological and physical well-being. Overall, the findings support the construct validity of both the 2-item measure and 2-item combined measure.

See extended abstract

 Presented in Session 9. Marriage, Family, Households, & Unions; Gender, Race, & Ethnicity