Abstract
The aim of the present study was to examine the dimensional structure of the Gambling Attitudes and Beliefs Survey (GABS). The GABS was administered to a sample of 415 individuals with self-reported problem or pathological gambling who were taking part in two different treatment studies preregistered with the German Clinical Trials Register (DRKS00013888) and ClinicalTrials.gov (NCT03372226). Exploratory factor analyses revealed a three-factor structure. We labeled the factors sensation seeking/illusion of control, luck/gambler's fallacy, and attitude/emotions. Subsequent confirmatory factor analyses proved the three-factor model superior to the one-factor model proposed by the developers of the GABS. All dimensions were significantly correlated with symptom severity scores. Group comparisons showed significantly higher factor scores on the first factor (sensation seeking/illusion of control) for individuals reporting both skill-based and chance-based gambling compared to those reporting only chance-based gambling. The present study questions the unidimensionality of the GABS. A multidimensional assessment of gambling-related cognitive biases, beliefs, and positively valued attitudes may be useful in determining treatment outcomes and goals and in the development of novel interventions.