Locating gambling problems across a continuum of severity : Rasch analysis of the Quinte Longitudinal Study (QLS)

Addressing gambling problems across a continuum requires understanding of low severity problems, as well as severe levels of problem gambling or disorder. The aims of this study were thus to derive a map of how problematic gambling behaviours and harms are situated across a continuum, and identify the best available indicators of low severity problems to inform assessment and secondary prevention. This involved the Rasch analyses of baseline data from the Quinte Longitudinal Study (QLS); a community-based survey involving random-digit dialling of numbers around Belleville, Canada. Participants were n = 1305 adults with non-zero scores across 26-items from: the Problem Gambling Severity Index (PGSI); the NORC DSM Screen for Gambling Problems (NODS); and the Problem and Pathological Gambling Measure (PPGM). Results indicated that item-level measures except chasing losses provided fit to the Rasch model, and most were clustered within a narrow region of the continuum which resembled addictive disorders. At the most severe end were mainly items about harms, while there were few items representing low severity levels (feeling guilty, betting more than one can afford, attempts to reduce gambling, gambling more than intended). There was Differential Item Functioning (DIF) for several indicators of low severity problems. The findings suggest that measures remain closely aligned with psychiatric models and are suited for discriminating across severe levels of problem gambling or addictive disorder. Although cognitive-affective and behavioural indicators comprise the best available indicators of low severity symptoms, there is an urgent need for improvements in conceptualisation and measurement.