An Item Response Theory-Based Scoring of the South Oaks Gambling Screen–Revised Adolescents

Abstract

The South Oaks Gambling Screen–Revised Adolescent (SOGS-RA) is one of the most widely used screening tools for problem gambling among adolescents. In this study, item response theory was used for computing measures of problem gambling severity that took into account how much information the endorsed items provided about the presence of problem gambling. A zero-inflated mixture two-parameter logistic model was estimated on the responses of 4,404 adolescents to the South Oaks Gambling Screen–Revised Adolescent to compute the difficulty and discrimination of each item, and the problem gambling severity level (θ score) of each respondent. Receiver operating characteristic curve analysis was used to identify the cutoff on the θ scores that best distinguished daily and nondaily gamblers. This cutoff outperformed the common cutoff defined on the sum scores in identifying daily gamblers but fell behind it in identifying nondaily gamblers. When screening adolescents to be subjected to further investigations, the cutoff on the θ scores must be preferred to that on the sum scores.

Online Gambling Disorder Questionnaire (OGD-Q): An item r‘esponse theory examination

“…The current findings are in line with the Sharp et al (2012) IRT of the PGSI, which showed that chasing one’s losses had the lowest discriminating ability compared to other items (α = 1.07). However, the SOGS-RA (which did not have a similar “tolerance” item), was found to perform moderately well at discriminating between individual responses with a yes/no response format when participants were asked about chasing their losses (λ = 0.65, α = 1.61; Anselmi et al, 2021 ).…”

“…In the SOGS-RA IRT study, the item that appears to be similar to OGD-Q item 11, (i.e., “Have you ever skipped or been absent from school or work due to betting activities?”) had the second highest difficulty level (β = 1.86; Anselmi et al, 2021 ), indicating that a moderate amount of the trait would be required to endorse “yes” to the item. However, it may not represent the same aspect of problem gambling as the OGD-Q item 11.…”

“…The PGSI does not have an item that aligns with prioritising gambling over other important functions, so a further comparison cannot be made. In relation to the second-highest ranked OGD-Q item for difficulty (i.e., item 7), the SOGS-RA has one item that may also conceptualise perceived lack of control (e.g., “Have you felt that you would like to stop betting money but didn’t think you could?”), which indicated that a moderate level of trait would be needed to endorse “yes” to the item (β = 1.71; Anselmi et al, 2021 ).…”

“…Other IRT studies have demonstrated that different components/criteria of behavioural addiction function differently psychometrically when operating as instruments in diagnostic scales, such as in food addiction ( Kircaburun et al, 2020 ), and internet gaming disorder ( Gomez et al, 2019 ), as well as problem gambling ( Anselmi et al, 2021 , Sharp et al, 2012 ). However, past IRT studies emphasizing on gambling scales to assess the differential psychometric functioning of disordered gambling criteria/ behaviours invite further investigation due to: a) in some cases exclusively examining adolescent and or specific national populations (e.g.…”

“…However, past IRT studies emphasizing on gambling scales to assess the differential psychometric functioning of disordered gambling criteria/ behaviours invite further investigation due to: a) in some cases exclusively examining adolescent and or specific national populations (e.g. South Africans) and; b) not reporting/examining pseudo-guessing indices ( Anselmi et al, 2021 , Sharp et al, 2012 ). Therefore, examining the OGD-Q using and advanced three parametric (3-PL) IRT analysis model seems promising, as it could determine a) which items (i.e., components) are more likely to be endorsed based on differing trait levels; b) which are more reliable or precise at measuring disordered gambling behaviours and; c) how pseudo-guessing effects could differentially effect one’s assessment.…”