Testing measurement invariance of the protective behavioral strategies scale in college men and women.

This study tested whether the measurement parameters of the Protective Behavioral Strategies Scale (PBSS; Martens et al., 2005) were equivalent for men and women in a college student sample. Multi-group confirmatory factor analyses tested whether a similar 3-factor latent structure was shared across gender groups and whether item origins (i.e., thresholds) and the strength of associations between individual items and latent factors (i.e., factor loadings) were equivalent across gender groups. Multiple indicators multiple causes (MIMIC) models tested whether gender differences in measurement parameters were explained by drinking patterns among college men and women. Results evidenced significant measurement bias (i.e., non-invariant measurement parameters) across gender groups at both the factor structure and individual item level. In addition, MIMIC models suggested that gender differences in item loadings and thresholds were not better explained by discrepancies in drinking patterns among male and female college students. Findings indicate that gender differences in latent factor scores may reflect measurement bias rather than true mean differences between gender groups, restricting meaningful comparisons of PBSS scores between college men and women. Implications for the assessment of specific strategies as well as clinical interventions among college students that include components on protective behavioral strategies are discussed. 2014 APA