The Influence of Guilt on Human Online Prosocial Behavior

Abstract

The nearing possibility of mixed societies involving both humans and (humanoid) robots, increases the importance of investigating the possible differences in behavior that people show towards robots and humans. Research on this topic is scarce and found to be inconsistent; it is unclear whether people treat robots differently than how they treat other people. As guilt is suggested to be a strong predictor of prosocial behavior, differences in feelings of guilt towards robots and humans were thought to cause differences in how they are treated. The aim of the current study was to investigate the relationship between guilt and human prosocial behavior and make a comparison between human-robot interactions (HRI) and human-human interactions (HHI). Data from 73 participants was used to investigate whether; (H1), participants interacting with another human displayed more prosocial behavior towards their opponent; (H2), participants experiencing guilt engaged in more prosocial behavior, regardless of type of opponent and if, (H3), more guilt is experienced in HHI compared to HRI. An online experiment was designed where participants interacted with an agent (human or robot) in a manipulated context (Yes lie or No lie). Guilt was manipulated by asking participants to lie to their opponent (Yes lie). A modified version of the Prisoner’s dilemma was used as a measure for prosocial behavior. Participants reported amount of guilt and empathy, and their perception of their opponent in additional questionnaires. The results were not conclusive; no evidence was found for the expected results and only partial evidence was found for the guilt manipulation. Scores on empathic concern could have suggested evidence of guilt, however further research is needed to substantiate this implication. For future research it would be interesting to conduct a similar study using face-to-face interaction, monitor the potential effect of cognitive dissonance and emphasize the role of empathy in human prosocial behavior. The present study has contributed to the fields of human emotion, human behavior and human-robot interactions by providing new insights for future research as it is important to not overlook the potential differences in the way we treat and interact with robots. This way, potential problems that may arise from the introduction of (humanoid) robots in societies, could be foreseen and prevented.