Validation of the 7-Item Domain-General Gambling Harm Scale (DGHS-7)
Journal article, Peer reviewed
Published version
![Thumbnail](/bora-xmlui/bitstream/handle/11250/3118186/1-s2.0-S2352853223000214-main.pdf.jpg?sequence=6&isAllowed=y)
View/ Open
Date
2023Metadata
Show full item recordCollections
- Department of Psychosocial Science [847]
- Registrations from Cristin [10237]
Abstract
Introduction
Gambling can cause negative consequences affecting finances, work/study, physical and mental health, relationships, law abidingness, and the community. Although existing measures enable investigations of gambling harms, there is still a need for a brief measure covering the full range of gambling related harms.
Methods
We validated a 7-item domain-general harm scale (DGHS-7) using data from a cross-sectional survey of United Kingdom residents reporting gambling within the last 12 months (n = 2558, 62.4% women, mean age 40.1 years (SD = 12.5)). The DGHS-7 was investigated in terms of factor structure, measurement invariance, and convergent validity with a comprehensive 72-item checklist of gambling harm, the Short Gambling Harms Screen (SGHS), and the Problem Gambling Severity Index (PGSI). Discriminative validity was checked against the Personal Wellbeing Index (PWI). Internal consistency was also calculated.
Results
Confirmatory factor analysis supported a one-factor solution (χ2 = 136.991, df = 14, χ2/df = 9.785, p <.001, CFI = 0.999, RMSEA = 0.059, 90% CI [0.050, 0.068]). Measurement invariance was supported for gender and binary categorization of age and income (ΔCFI = 0.001). The DGHS-7 correlated strongly with the 72-item checklist (rs = 0.824), the SGHS (rs = 0.793), the PGSI (rs = 0.768), and moderately with the PWI (rs = -0.303). Cronbach’s alpha = 0.91 and ordinal alpha = 0.96 indicated good internal consistency.
Conclusions
Psychometric support was found for a brief measure covering all recognized domains of gambling harm. The DGHS-7 is useful for researchers needing a generic and short measure for epidemiological and other studies calling for short scales.