Agnieszka Tymula (Sydney)
Behavioural and neural evidence on probability weighting
We use a large dataset of lottery choices made by laboratory monkeys to estimate the static and dynamic distortions in probability. Our new model, incorporating trial-by-trial prediction-error dynamics into probability perception, best describes an animal’s subjective belief of winning. We confirm similar patterns in human behaviour under similar laboratory conditions. We then link the behavioural data to neural measurements made throughout the reward circuitry in the monkey brain. We propose a network model that aggregates these signals and reliably reconstructs risk preferences and subjective probability perceptions revealed by the animals’ choices.