Factorial invariance of the Everyday Discrimination Scale among African, Caribbean, Black, and other racialized women living with and without HIV in British Columbia, Canada.

Objectives: The Everyday Discrimination Scale (EDS) assesses the mistreatment racialized people experience daily. It is widely used in Canadian HIV research to measure racial discrimination among African, Caribbean, and/or Black (ACB), and other racialized women. However, no studies have examined whether the EDS provides valid estimates of racial discrimination that are objectively comparable across racial/ethnic groups or by HIV status in Canada. We aimed to address this gap.

Methods: We analyzed EDS data from the British Columbia CARMA-CHIWOS Collaboration study of self-identifying women ≥ 16 years of age living with and without HIV. We conducted confirmatory factor analysis (CFA) by race/ethnicity (ACB; Indigenous; other racialized; white) and also by HIV status to assess validity and reliability of the EDS within these groups. Next, we ran multigroup-CFA (mCFA) to determine whether the EDS measures the same construct to the same extent across race/ethnicity- and HIV status-based groups.

Results: Of 504 participants (202 with HIV), 62 were ACB, 152 Indigenous, 90 other racialized, and 200 white. CFA revealed adequate fit within race/ethnicity- and HIV status-based groups-all Cronbach's alpha > 0.9, factor loadings > 0.6, and acceptable fit indices, except for root mean square error of approximation. Residual correlations > 0.4 suggested redundancy between some item pairs. In the mCFA, the EDS was invariant across groups.

Conclusion: The EDS provides racial discrimination estimates that are meaningfully comparable across race/ethnicity- and HIV status-based groups. Despite potential redundancy between item pairs that may warrant refinement, the EDS can be included in survey instruments that aim to examine and compare the prevalence and health impacts of racial discrimination among ACB and other racialized women in Canada.