Can Serious Games Assess Decision-Making Biases? Comparing Gaming Performance, Questionnaires, and Interviews

The limitations of self-report questionnaires and interview methods for assessing individual differences in human cognitive biases have become increasingly apparent. These limitations have led to a renewed interest in alternative modes of assessment, including for implicit and explicit aspects of human behavior (i.e., dual-process theory). Acknowledging this, the present study was conducted to develop and validate a serious game, "Don Quixote," for measuring specific cognitive biases: the bandwagon effect and optimism bias. We hypothesized that the implicit and explicit game data would mirror the results from an interview and questionnaire, respectively. To examine this hypothesis, participants (n = 135) played the serious game and completed a questionnaire and interview in a random order for cross-validation. The results demonstrated that the implicit game data (e.g., response time) were highly correlated with the interview data. On the contrary, the explicit game data (e.g., game score) were comparable to the results from the questionnaire. These findings suggest that the serious game and the underlying intrinsic nature of its game mechanics (i.e., evoking instant responses under time pressure) are of importance for the further development of cognitive bias measures in both academia and practice.