Learning from social rewards predicts individual differences in self-reported social ability.

The ability to adapt face-to-face social behavior in response to an interaction's changing contingencies is an important aspect of social skill. Individual differences in social ability may depend on how well people learn from social rewards and punishments. Here we relate people's social aptitude to their ability to learn from differences in the reward values of two common social reinforcers, genuine and polite smiles. In a series of experiments, participants experienced a hidden social contingency in which they either learned to repeat actions that received genuine smile feedback and switch after polite smiles or the reverse. A condition with nonsocial feedback served as a comparison measure. Participants showed better ability to repeat actions reinforced with genuine smile feedback than with nonsocial feedback. When participants were required to switch actions following genuine smiles, performance was inhibited relative to nonsocial reinforcement. The ability to detect task contingencies and learn from social rewards predicted self-reported social ability. These novel results suggest that individual differences in reinforcement learning, and particularly in people's motivation to receive social rewards, may relate to social ability in face-to-face interactions. This finding has important implications for understanding the social difficulties that characterize disorders such as autism, depression, and schizophrenia, in which the ability to learn from rewards may be compromised.