Investigating the nature of children's altruism using a social humanoid robot

Abstract Young children help people achieve their goals in various situations, even in the absence of verbal requests or rewards. Such help is typically described as “indiscriminate” because children's help is robust across many aspects of the recipient in need. However, as previous studies have almost entirely focused on human recipients, the question arises as to how indiscriminate children's helping behaviour really is. Here, we aimed to investigate whether young children's helping behaviour would extend to a nonhuman entity—a humanoid robot. In an experimental between-subjects design, 40 children witnessed a robot drop an object out of its reach either seemingly intentionally or unintentionally. Results showed that children typically returned the object, but only when it was dropped apparently unintentionally—indicating that children attribute goals to a robot, understand when it requires help, and are motivated to help it. Results of a post-test animacy interview were consistent with this interpretation as children attributed affective, cognitive, and physiological characteristics to the robot. We conclude that the indiscriminate nature of young children's altruistic behaviour extends to humanoid robots. This finding has important theoretical and practical implications for the future development of social robots as well as for the study of children's prosocial behaviour.