Validity of the Problem Gambling Severity Index interpretive categories

Area of Science:

• Psychology
• Public Health
• Behavioral Science

Background:

• The Problem Gambling Severity Index (PGSI) categorizes gamblers into non-problem, low-risk, moderate-risk, and problem types.
• Existing validity testing primarily focused on problem gamblers, neglecting the majority (over 95%) of gamblers in other categories.

Purpose of the Study:

• To conduct a comprehensive validity and reliability analysis of all four PGSI gambler types.
• To assess the temporal stability of PGSI gambler classifications.
• To improve the discriminant validity of the low-risk and moderate-risk categories.

Main Methods:

• Analysis of Canadian population data from over 25,000 gamblers.
• Assessment of temporal stability over a 14-month interval using intraclass correlation coefficients.
• Evaluation of discriminant validity for all PGSI categories with existing and modified scoring rules.

Main Results:

• Modest but adequate temporal stability for PGSI subtypes (ICC = 0.63).
• Strong evidence for the validity of non-problem and problem gambler categories.
• Poor discriminant validity for low-risk and moderate-risk categories under existing scoring rules.
• Improved validity for low-risk and moderate-risk categories following a scoring system modification.

Conclusions:

• The PGSI demonstrates adequate reliability and validity for non-problem and problem gambler types.
• The validity of low-risk and moderate-risk categories is significantly enhanced by a revised scoring system.
• The findings support the utility of a modified PGSI for more accurate gambling severity assessment across all gambler types.