People often exhibit more socially favorable behaviors when observed by others, potentially influencing their cognitive skills and prosocial tendencies. Recent studies have found that individuals with intrinsic prosocial tendencies are non-responsive to social observation in various prosocial decision tasks. This study aimed to investigate whether individuals with intrinsic prosocial tendencies also exhibit a lack of change in their cognitive ability under social observation. We used the Prosocial Reinforcement Learning Task (PRLT) to assess the interaction effect of social observation and intrinsic prosocial tendency on prosocial learning tendency. A total of 102 participants were randomly assigned to either the observation or control group while performing a two-armed bandit task under self- and other-reward conditions, and their behavioral outcomes were analyzed using a reinforcement learning computational model. Under social observation, participants who were previously less prosocial exhibited increased prosocial learning. In contrast, those who were already more prosocial showed no significant changes in prosociality, and demonstrated only a numerical-but statistically non-significant-increase in learning for self. Our findings revealed the differential effects of social observation on modulating one's prosociality and cognitive ability according to individual differences in intrinsic prosocial tendencies.
