How different are real and hypothetical decisions? Overestimation, contrast and assimilation in social interaction

Abstract

Decision researchers conduct laboratory experiments of choice between real rewards, but also often rely on asking people to provide hypothetical answers to fictitious situations. Applying results from such studies to real-world situations requires understanding how decision making in cases involving real rewards correspond to decisions with hypothetical rewards. This article reports two key differences between hypothetical and real decisions in social dilemmas. First, hypothetical and real decisions cause different cognitive biases in social dilemmas: hypothetical judgements and choices stem from perceptual processes and cause contrast effects; while judgements and actions in real interaction are biased by assimilation processes involved in action selection and learning. Second, without the corrective real social interaction, people overestimate theirs and others propensity to act cooperatively. However, individuals are more confident in their predictions during real interaction, which suggests that such metacognitive judgments can predict actual behaviour and also signal when respondents provide biased responses. Overall, our results call into question established methodologies that rely on hypothetical answers, and indicate that people should be observed and measured in real or incentivised social interactions.