Rapid changes in risk attitudes originate from Bayesian inference on parietal magnitude representations

Abstract

Risk attitudes—the willingness to accept uncertainty for larger potential rewards—are often treated as a stable individual trait, yet people’s risky choices change markedly across contexts and even across repetitions of identical decisions. Here we show, using modelling and neuroimaging, that such rapid changes emerge from Bayesian inference on noisy magnitude representations in parietal cortex. Participants underwent fMRI while choosing between safe and risky options that are either held in working memory or shown on screen. Options held in working memory are noisier and more prone to central-tendency biases, so relatively small payoffs are overestimated and large payoffs underestimated, shifting choices between risk-seeking and risk aversion. A computational model of payoff perception captures these effects mechanistically. Numerical population-receptive-field modelling shows that when intraparietal payoff representations are noisier, choices are less consistent and less risk-neutral. Thus, risk attitudes and their contextual variability can be rooted in perceptual inference on parietal magnitude representations.