When a Robot's Group Membership Matters

Previous work has documented that humans categorize robots as members of different social groups, thereby using socially relevant cues such as a robot's alleged gender or nationality. Importantly, these social categorization processes affect impressions people form about robots. In an experiment with N = 45 participants, we utilized a minimal-group paradigm and tested whether categorizing the humanoid robot NAO as an in-group member vs. an out-group member based on socially non-relevant features would result in higher levels of anthropomorphism and more positive evaluations of the robot. Innovatively, to assess anthropomorphism, we utilized an implicit measurement procedure. Results support our hypotheses: Perceived in-group membership with the robot resulted in a greater extent of anthropomorphic inferences about the robot and more positive evaluations. Moreover, compared to the out-group condition, participants who perceived NAO as an in-group member showed greater willingness to interact with robots in general.