Measurement Invariance of the World Assumptions Questionnaire across Race/Ethnic Group, Sex, and Sexual Orientation

Document Type


Publication Date



Objective: The World Assumptions Questionnaire (WAQ) was developed to assess optimism and assumptions about the world, which often shift after traumatic events. However, no known study has investigated whether the WAQ holds similar meaning across demographic groups. The objective of this study was to investigate measurement invariance of the WAQ across race/ethnic group, sex, and sexual orientation. Method: Participants consisted of 1,181 college students (75% female; 25% Black, 13% Latinx, 18% Asian, 45% White; 90% heterosexual) who completed an online survey on stress, personality, substance use, and mental health. We investigated a unidimensional and the 4-factor structure of the WAQ using confirmatory factor analysis, and configural, metric, and scalar invariance using multigroup confirmatory factor analysis. Results: After dropping 3 items, a 4-factor structure fit the data well (comparative fit index = .92; root mean square error of approximation =.05; 95% confidence interval [.045, .054]; standardized root mean square residual = .06). Mean WAQ scores were higher for participants with probable posttraumatic stress disorder on 2 of the 4 factors. We also identified multiple items that were not invariant across race/ethnic group, sex, and sexual orientation. However, after invariant items were removed, evidence of configural, scalar, and metric invariance was found. Conclusions: This study replicated the 4-factor structure, mapping onto the 4 WAQ subscales, and indicated that a unidimensional measure of world assumptions should not be used. After making the adjustments recommended herein, the WAQ can be used to investigate differences across race/ethnic group, sex, and sexual orientation.