A Novel Game-Based Assessment of Risk-Taking Propensity: Development and Psychometric Evidence

Risk-taking propensity is defined and measured in different ways across the literature, resulting in assessment issues related to construct validity (Bran & Vaidis, 2020). Multimethod assessment approaches can bolster construct validity by capturing more facets of risk-taking propensity (Meyer et al., 2001). Risk-taking propensity is assessed through self-report questionnaires (Zhang et al., 2019), which exhibit good test-retest reliability (Frey et al., 2017) and predictive validity (Beauchamp et al., 2017), and behavioral tasks (Aklin et al., 2005), which are less psychometrically sound, lacking convergent validity and test-retest reliability (Frey et al., 2017). Guided by evidence-centered design principles (Mislevy, 2013), a novel game-based assessment of risk-taking propensity may mitigate the current shortcomings of behavioral tasks. Exploratory factor analysis revealed a well-fitting model and good reliability of the in-game instrumental risk-taking items. The in-game scores weakly correlated with the self-report measure of instrumental risk-taking, but did not have significant correlations with a set of related variables (e.g., financial risk-taking or general risk-taking propensity). The in game instrumental risk-taking scores were predictive of real-world risk-taking behaviors, both in isolation and when modeled with the self-report scores. The in-game assessment is tenable for use in a multimethod approach of measurement in research designs. With dual consideration of both assessment design and game design, game-based assessments can be developed and validated for use in future research.