Measurement and Structural Invariance of Posttraumatic Stress Disorder Symptoms in Hispanic and Caucasian Firefighters: A Bias-Corrected Bootstrap Confidence Intervals Approach

Abstract

This study compared across Hispanic and Caucasian firefighters the relative fit of the four-factor Emotional Numbing and Dysphoria posttraumatic stress disorder models to the more recently proposed Dysphoric Arousal five-factor model. As hypothesized, the Dysphoric Arousal five-factor model emerged as the best fitting model within each ethnic group and it also showed measurement invariance between groups (configural invariance). Results of multigroup confirmatory factor analyses and a bias-corrected bootstrap confidence intervals analytic approach indicated that the five factor model also demonstrated invariance in factor loadings (metric invariance) and item-level intercepts (scalar invariance) across the two ethnic groups. Results indicate that the Dysphoric Arousal five factor model captures similar psychological constructs across Caucasian and English-speaking Hispanic firefighters. Therefore, observed factor scores are comparable across ethnic groups and can be combined when examining predictors of posttraumatic stress disorder severity.