关键词: 社交机器人, 亲社会行为, 分享, 心理能动性

Abstract: Prosocial behavior benefits both individuals and society. Previous research has shown that human observers can influence the prosocial behavior of 5-year-old children. Psychological agency refers to the ability to think, make decisions, possess knowledge reserves, and interact responsively. Prior studies have indicated that the psychological agency of robots affects children's selective trust in them. Based on this background, this study aimed to understand the impact of social robots on children's prosocial behavior in the context of human-robot interaction. Experiment 1 examined the effect of different observer types on preschool children's prosocial behavior, utilizing a 3 (observer type: human, social robot, none) × 2 (age group: 3 years, 5 years) between-subjects design. The experimenter had all participating children interact with a social robot before administering an animism test. Then, the children were randomly assigned to one of three groups: human observer, social robot observer, and no observer, and participated in an anonymous dictator game to share stickers. After the game, the experimenter asked and recorded the reasons for the children's sticker-sharing behavior. Experiment 2 further investigated the effect of different levels of robot psychological agency on children's prosocial behavior. In Experiment 2, 3-year-old children were recruited and randomly assigned to the groups with and without psychological agency of social robots, as well as a control group. Children in the psychological agency groups interacted with a social robot and then took an animism test. Subsequently, the children participated in the anonymous dictator game in front of the robot. The control group children did not interact with the robot and directly participated in the anonymous dictator game. After each group of sharing games, the experimenter interviewed and recorded the reasons for the children's sticker-sharing behavior. In both experiments, the experimenter recorded the number of stickers shared by the children after they left. The study found that 5-year-old children shared more stickers and exhibited more prosocial behavior as compared to 3-year-old children. For 3-year-old children, the number of stickers shared in the human and robot observer conditions was significantly higher than in the no observer condition, with no significant difference between the human and robot observer conditions. Additionally, the study manipulated the psychological agency of social robots to explore its effect on the prosocial behavior of 3-year-old children. The results showed that 3-year-old children shared more stickers in the presence of a robot with psychological agency compared to the other two conditions. The study also explored the animism tendencies of children toward robots, indicating that 5-year-old children had exhibited less animism than 3-year-old children. Furthermore, 3-year-old children attributed more psychological and life-like characteristics to robots with psychological agency as compared to those without. The research confirms that observers influence children's prosocial behavior and finds that the prosocial behavior of 3-year-old children is affected by the psychological agency of robots. Additionally, this study explores the animism tendencies of Chinese children of different ages toward social robots. These findings provide perspectives and foundations for future applications and research in the field of human-robot interaction.

Key words: Social robots, Prosocial behavior, Sharing, Psychological agency