Cross Validation of the Gambling Problem Severity Subscale of the Canadian Adolescent Gambling Index (CAGI/GPSS) on a Sample of Ontario High School Students

Nigel E Turner,Tara Elton-Marshall,Jamie Wiebe,Angela Boak,Robert E Mann,Jing Shi,Mark Van Der Maas

doi:10.1007/s10899-017-9731-1

Abstract

This paper reports on the cross validation of the Gambling Problem Severity Subscale of the Canadian Adolescent Gambling Index (CAGI/GPSS). The CAGI/GPSS was included in a large school based drug use and health survey conducted in 2015. Data from students in grades 9–12 (ages 13–20 years) derived from the (N = 3369 students). The CAGI/GPSS produced an alpha of 0.789. A principle component analysis revealed two eigenvalues greater than one. An oblique rotation revealed these components to represent consequences and over involvement. The CAGI/GPSS indicated that 1% of the students fell into the “red” category indicating a severe problem and an additional 3.3% scored in the “yellow” category indicating low to moderate problems. The CAGI/GPSS was shown to be significantly correlated with gambling frequency (r = 0.36), largest expenditure (r = 0.37), sex (more likely to be male) (r = −0.19), lower school marks (r = −0.07), hazardous drinking, (r = 0.16), problem video game play (r = 0.16), as well as substance abuse. The CAGI/GPSS was cross validated using a shorted version of the short SOGS, r = 0.48. In addition the CAGI/GPSS and short SOGS produced very similar patterns of correlations results. The results support the validity and reliability of the CAGI/GPSS as a measure of gambling problems among adolescents.

Highlights

Adolescents are often viewed as being vulnerable to gambling-related harms (Gupta and Derevensky 1998a; Shaffer et al 1999; Turner et al 2008a)
This paper reports on the cross validation of the Gambling Problem Severity Subscale of the Canadian Adolescent Gambling Index (CAGI/GPSS)
We examined the relationship between the CAGI/GPSS and the short SOGS

Summary

Introduction

Adolescents are often viewed as being vulnerable to gambling-related harms (Gupta and Derevensky 1998a; Shaffer et al 1999; Turner et al 2008a). Some studies of the prevalence of problem gambling among adolescents have found the rate is two to three times higher than that of adults (Gupta and Derevensky 1998a; Ladouceur 1996; Shaffer and Hall 2001; Shaffer et al 1999; Turner et al 2008a; Wiebe et al 2005). Wiebe et al (2005) noted that the estimated prevalence varies from study to study in part because of differences in the measures and methods used These previous measures were either created for adult with gambling problems or were adapted for adolescents (Wiebe et al 2005).

Methods

Results

Discussion

Conclusion