Children’s psychological and moral attributions to a humanoid robot

In the near future, sophisticated social robots will become increasingly interwoven into our lives. Researchers have recently begun to examine people’s anthropomorphic conceptions of such robots, and a few have stressed the unique consequences that these technological agents may have for the psychological development of children developing around them. In the current set of studies, children were introduced to a humanoid robot, “Robbie the Robot.” Across the two studies, participants witnessed Robbie perform a harmful action, destroying a block tower that a child had purportedly built and was saving for later. Of primary interest in these two studies was whether children would hold Robbie the Robot morally accountable for the destructive act. It was predicted that judgments of moral accountability would depend on several different factors: whether the robot appeared to initiate its own actions, the age of the participant, and whether children attributed psychological properties, specifically intentional agency, to the robot. In Study 1, children were assigned to one of two experimental conditions: a controlled condition in which a confederate appeared to control the robot’s actions with a device that was tethered to the robot, and an autonomous condition in which the robot appeared to move of its own accord. Results revealed that children were significantly more likely to attribute psychological properties to the robot in the autonomous condition compared to the controlled condition. Compared to 7-year-olds, 5-year-olds were more likely to attribute psychological properties to the robot overall. In addition, results indicated that increasing cues to the robot’s autonomy indirectly affected moral accountability judgments through an increase in children’s attributions of intentions. Study 2 tested the hypothesis that children’s attributions of psychological agency, but not psychological experience, would increase after watching the robot commit a moral act. Overall, Study 2 results did not support this prediction, but key results from the first study were replicated and elucidated by the inclusion of a wider array of psychological properties as well as a measure of children’s judgments of the robot’s cuteness. Implications are discussed for human interaction with social robots and other rapidly evolving technologies, such as autonomous vehicles.