Construct validity of the Short Warwick-Edinburgh Mental Well-Being Scale in a diverse urban population

Purpose: Although the Short Warwick-Edinburgh Mental Well-Being Scale (SWEMWBS) has been validated in samples across several countries it has yet to be validated in the US. This study examines the validity of the SWEMWBS within a diverse urban population of US adults. Methods: A population representative sample of adults living in NYC was obtained from the NYC Neighborhood Wellness Survey (NWS). Respondents answered questions about their mental health, emotional and social support, demographics (i.e., age, sex at birth, gender identity, sexual orientation, and race/ethnicity) and completed the SWEMWBS. Using a combination of Rasch analyses and classical test theory (CTT), we assessed the SWEMWBS for unidimensionality, monotonicity, local independence, differential item functioning, and measurement invariance. Results: Ninety-five percent of eligible participants completed all seven items of the SWEMWBS (N=41,484). The measure showed high internal consistency (α = 0.88), and all items were highly correlated with the overall score. Assumptions of unidimensionality, monotonicity, and local independence were supported. Differential item functioning was identified between age groups (18-24 vs 65-74; 18-24 vs 75+; 25-29 vs 65-74; 25-29 vs 75+; 30-44 vs 75+) for SWEMWBS item 1 (I've been feeling optimistic about the future). Measurement invariance was assessed for key demographic variables, but no significant differences between configural and metric models were identified. Conclusion: Despite differences in item functionality by age, this assessment shows that the SWEMWBS is a useful tool for evaluating mental well-being in a diverse population and should be studied further among smaller groups and rural communities in the US.