Work

Measurement invariance of the PTSD checklist for the DSM-5 (PCL-5) in a veteran sample.

Year: 2025

Type: article

Abstract: = 1,357, 51% women, 18% Black) of post-9/11 Army and Marine Corps veterans overselected for PTSD symptoms. PTSD was modeled within a higher order factor structure using observed PCL-5 responses as ordered polytomous data. We examined group-level invariance across item response thresholds and first- and second-order factor loadings and intercepts. We used 500 bootstrapped racially balanced data sets to account for sample size imbalance in the racial invariance analyses. Results provide support for partial measurement invariance across gender with veteran men reporting significantly more hypervigilance and reckless and self-destructive behavior than women at equal levels of latent PTSD symptom severity. Importantly, we found full support for invariance across Black and White veterans. These results suggest that PCL-5 score differences observed across these groups represent substantive differences in PTSD symptoms rather than bias associated with the measurement process. (PsycInfo Database Record (c) 2025 APA, all rights reserved).

Source: Psychological Assessment

DOI: 10.1037/pas0001396

Authors: Michael L. Crowe, Nicholas A. Livingston, Terence M. Keane, Brian P. Marx

Institution: VA Boston Healthcare System

Language: English

Topic: Psychometric Methodologies and Testing

Subfield: Management Science and Operations Research

Field: Decision Sciences

Domain: Social Sciences

SDGs: Good health and well-being, Reduced inequalities

Open Access status: hybrid

Funders: U.S. Department of Defense, U.S. Department of Veterans Affairs