Reward Responsiveness and Inhibition Traits Differentially Predict Economic Biases in Gain and Loss Contexts.

Abstract

Research on economic decision making has revealed specific biases in gain versus loss domains such that risky choice options are overvalued in gain conditions, implying optimism, but undervalued in loss conditions, implying pessimism. Individual differences in motivational traits and affective states have been shown to predict beliefs and behavior in risky decision making, but it is presently unclear which personal characteristics are most predictive of domain-specific biases. To address this gap in the literature, we investigated the relative influence of positive and negative motivational traits (general sensitivity to rewards and punishments) versus affective states (current levels of positive and negative emotions) on beliefs and choice behavior during a risky economic decision task. We also expanded on previous research by examining how the valence of one’s judgment context (positive context tested in Experiment 1, negative context tested in Experiment 2) may determine whether risky choice behavior is more strongly influenced by positive versus negative characteristics. Biases in belief were calculated using an economic decision task that involved estimating the value of risky “stocks” relative to safe “bonds” from experienced outcomes. Experiment 1 used a positive judgment context (likelihood of a “good stock”) while Experiment 2 used a negative judgment context (likelihood of a “bad stock”). Consistent with previous findings, we observed a domain-based bias in beliefs about stock values across experiments, such that participants exhibited optimism in gain domain and pessimism in the loss domain. Experiment 1 further revealed that domain-based bias and suboptimal choice behavior was predicted by trait-level reward sensitivity, while positive affective state (PAS) had a more limited influence on belief bias alone. Under the negative judgment context of Experiment 2, there was a similar relationship between reward sensitivity and choice behavior; however, results revealed a slightly stronger influence of negative affective state (NAS). A subsequent cross-study analysis found sensitivity to rewards was most predictive of domain-based biases. These results suggest that motivational traits – particularly those relating to reward sensitivity – are more consistent predictors of domain-based biases and risky choice behavior than affective states, but their predictive power depends the valence of the decision context.