# Measuring Perceived Discrimination and Its Consequences for Latino Health

## TL;DR

This study shows that how discrimination is measured affects health outcomes, and better measurement can reveal stronger links between discrimination and health issues among Latino groups.

## Contribution

The study introduces a multidimensional approach to measuring discrimination that reveals stronger health associations and ethnic differences within the Latino population.

## Key findings

- Significant ethnic differences in perceived discrimination were found among Latino groups.
- A second-order factor model of discrimination (subtle and overt) fits better than unidimensional scales.
- Modeling discrimination as a second-order factor strengthens its link to health outcomes like depression and chronic conditions.

## Abstract

Research demonstrates that discrimination is detrimental to health. However, most discrimination research does not examine Latino ethnic differences and often relies on unidimensional alpha scales. Such an analytic strategy obscures ethnic differences, can mask the multidimensional nature of discrimination, inflate reliability estimates, produce attenuated or spurious relationships, and bias parameters. To address these issues, we use data from the National Latino and Asian American Study to (1) examine group differences on the Everyday Discrimination Scale (EDS), (2) conduct a confirmatory factor analysis of the EDS to assess its fit and dimensionality for each Latino ethnic group, and (3) evaluate how alternative scaling approaches affect the relationship between discrimination, depression, and chronic health conditions. Results reveal significant group differences in perceived discrimination and show that a second-order factor with two dimensions—subtle and overt discrimination—fits well across all Latino groups. The relationship between discrimination and health is stronger when discrimination is modeled as a second-order factor. These findings indicate that (1) alternative scaling approaches may be more appropriate than alpha scales, (2) more precise measurement of discrimination can better capture its impact on health, and (3) disaggregating panethnic categories such as “Latino” that is essential for understanding ethnic stratification and health.