Examining the Factor Structure of the Substance Use Risk Profile Scale (SURPS) in Emerging Adults: An Exploratory Structural Equation Modelling Approach

Abstract

The Substance Use Risk Profile Scale (SURPS) measures personality traits linked to heavy drinking and related problems (hopelessness, anxiety sensitivity, impulsivity, sensation seeking) and informs personality-matching interventions. The SURPS’ factor structure shows inconsistencies, and evidence suggests that confirmatory factor analysis (CFA) is too restrictive for measures capturing correlated constructs. We examined if exploratory structural equation modelling (ESEM) better captured the optimal SURPS factor structure in a large Canadian sample, tested measurement invariance across sex and alcohol use differences, and assessed the predictive validity of SURPS subscales for alcohol use motives and problems. A sample of 6,397 emerging adults completed surveys. ESEM had excellent fit relative to CFA; Item 22 was removed due to a poor factor loading. The final model was invariant across groups; SURPS subscales predicted alcohol use motives and problems. Results support the SURPS’ utility for measuring substance use personality risk and ESEM’s utility for analyzing correlated constructs.