Anhedonia and Anxious Arousal Are Associated with Distinct Beliefs about the Statistics of a Volatile Environment

Using computational modeling, we showed that anhedonia and anxious arousal—key markers of depression and anxiety—involve diverging beliefs about the statistics of a volatile environment. Anhedonia predicted lower initial expectations of volatility across rewarding and aversive reversal-learning tasks, whereas anxiety predicted higher initial expected volatility specifically in aversive conditions.