Can social robots affect children's prosocial behavior? An experimental study on prosocial robot models

Abstract

The aim of this study was to investigate whether a social robot that models prosocial behavior (in terms of giving away stickers) influences the occurrence of prosocial behavior among children as well as the extent to which children behave prosocially. Additionally, we investigated whether the occurrence and extent of children's prosocial behavior changed when being repeated and whether the behavior modeled by the robot affected children's norms of prosocial behavior. In a one-factorial experiment (weakly prosocial robot vs. strongly prosocial robot), 61 children aged 8 to 10 and a social robot alternately played four rounds of a game against a computer and, after each round, could decide to give away stickers. Children who saw a strongly prosocial robot gave away more stickers than children who saw a weakly prosocial robot. A strongly prosocial robot also increased children's perception of how many other children engage in prosocial behavior (i.e., descriptive norms). The strongly prosocial robot affected the occurrence of prosocial behavior only in the first round, whereas its effect on the extent of children's prosocial behavior was most distinct in the last round. Our study suggests that the principles of social learning also apply to whether children learn prosocial behavior from robots.