Disordered gambling, or dependence and consequences: a bifactor exploratory structural equation model analysis of the Problem Gambling Severity Index

Objective: The Problem Gambling Severity Index (PGSI) is the most common population measurement of disordered gambling. However, there is some debate regarding the construct being measured, and whether it is a single factor of gambling problems, or separate indices of dependence and harms. The existing literature using exploratory and confirmatory factor analysis has a number of notable limitations that mean these accounts cannot be discriminated. Method: To overcome these, exploratory structural equation modelling was applied to secondary data from 10 nationally representative surveys of gamblers in the UK (n = 35,776) that were administered the PGSI, to test between these accounts. Results: Measurement indices provided slight support for a two-factor model. However, these models were not consistent across items or datasets, and further examination provided stronger support for a one-factor model. Sensitivity analyses coarsening across categories further supported these findings, whereas analysis looking at higher severity gamblers were not conclusive. Conclusions: This study provides strong evidence to support the use of PGSI scores as a single-factor construct. There are some circumstances (e.g., high engagement, clinical population) where a two-factor model may be appropriate to consider, but these require validation.