Evaluating lifetime NODS using Rasch modelling

Abstract

Rasch modelling was conducted by applying the lifetime NODS criteria using data from a national representative sample aged from 15 to 74 years in Norway (N = 5235). To a large extent, the results replicated previous findings, supporting the view that the DSM-IV gambling symptoms have a unidimensional structure. Differential item functioning (DIF) analysis indicated that younger gamblers (aged 15–25 years) were more likely to endorse the symptom of ‘chasing’ than older gamblers. Likewise, DIF analysis indicated that female gamblers were more likely to report ‘escape’ at lower levels of problem gambling than males. Moreover, the results showed that younger gamblers were less likely to endorse symptoms of ‘withdrawal’ and ‘loss of control’ than older gamblers. As the results may be explained by age and gender-specific correlates (or behaviour), future studies should include extended age groups (e.g. 12–75 years) or conduct Rasch modelling and DIF analysis on specific adolescent gambling instruments.