In reinforcement learning theory, humans learn from the outcomes of their actions and update the expected value of their future choices accordingly. To act in a socially adaptive manner, we must...Show moreIn reinforcement learning theory, humans learn from the outcomes of their actions and update the expected value of their future choices accordingly. To act in a socially adaptive manner, we must learn about the consequences our actions have on both ourselves and others. In the current study, empathy was tested as a trait which influences our ability to learn to make decisions which benefit others. It was hypothesised that higher empathy would lead to improved prosocial learning, and that feelings of responsibility for others would meditate such effect. A probabilistic prosocial reinforcement learning task was used, whereby 30 healthy males aged between 19 and 34 played a game to win monetary rewards for themselves, another person, or no-one. ANOVA analysis revealed that participants showed higher learning rates when playing for others rather than themselves, which is not congruent with previous research. The potential reasons for this finding are discussed. Correlation analysis of accuracy rates and computational learning rates with empathy scores revealed no relationships between trait empathy and prosocial learning. Further analysis failed to show feelings of responsibility for others mediating the effect of empathy on prosocial learning. Thus, the current study found no evidence for empathy having a role in prosocial learning, nor for feeling responsible for others as a mediator. The current sample did, however, perform as well or better for others than for themselves, which may be due to cultural differences or testing occurring during the Covid-19 pandemic when empathy, prosocial actions, and social responsibility were increased.Show less
Introduction The reinforcement learning theory shows that learning for another (prosocial learning) and learning for ourselves (selfish learning) can both be used as an effective way to learn. Low...Show moreIntroduction The reinforcement learning theory shows that learning for another (prosocial learning) and learning for ourselves (selfish learning) can both be used as an effective way to learn. Low self-esteem is linked to many clinical disorders and with prosocial and selfish behavior. The current study further examines selfish and prosocial learning and a possible relation with self-esteem. Method A total of 139 healthy participants finished the Rosenberg Self-esteem Scale and performed an online learning task. Participants had to choose between different stimuli that were probabilistically associated with rewards for themselves (self), another person (prosocial), or no one (control). The number of high probable stimuli (correctly chosen trials) for the selfish condition, prosocial condition and the none condition were analyzed. The two probability ratios that were used were 40/60 and 30/70. Results In contrast with our hypothesis, in the 30/70 probability context there was no significant difference found between conditions over time. Additionally, no significant effect was found between conditions and the 25% high and low self-esteem over time. Additionally, no significant correlations were found between self-esteem and the difference score ‘Prosocial minus Selfish’, the selfish condition and the prosocial condition. Discussion The current study found no significant relation between selfish and prosocial learning and self-esteem. The insignificant effects found in this study may be due to the difficulty of the task, uncontrollable environment, in diversity of sample and differences between global and specific self-esteem. Future research should focus on providing better insight in the relation between prosocial learning and self-esteem.Show less
