Abstract:The Iowa Gambling Task (IGT) is one of the most prominent paradigms employed for the assessment of risk taking in the laboratory, and it was shown to distinguish between various patient groups and controls. The present study was conducted to test the psychometric characteristics of the original IGT and of a new gambling task variant for assessing individual differences. Two studies were conducted with adults of the general population ( n = 220) and with adolescents ( n = 389). Participants were also tested on multiple measures of working memory capacity, fluid intelligence, personality traits associated with risk-taking behavior, and self-reported risk taking in various domains. Both gambling tasks had only moderate retest reliability within the same session. Moderate relations were obtained with cognitive ability. However, card selections in the gambling tasks were not correlated with personality or risk taking. These findings point to limitations of IGT type gambling tasks for the assessment of individual differences in risky decision making.

Psychometrics of the Iowa and Berlin Gambling Tasks: Unresolved Issues With Reliability and Validity for Risk Taking