Improving Human-Robot Interaction: Modifications of a Social Robot on Dimensions of Gender and Personality

Previous research in Human-Robot Interaction (HRI) has shown that social female robots typically evoked more cognitive and affective trust. Studies also show that extroverted robot personalities are associated with desirable social outcomes. In this experiment, I tested the extent to which a robot's gender (male v.s. female) and personality (introverted v.s. extroverted) impact the success of a given human-robot interaction. Specifically, I evaluated if the gender of a robot has an effect on the human preference for extroverted personalities. Prior to interacting with the robot, participants completed a baseline Negative Attitudes Towards Robots (NARS) scale. Then, during the interaction, the robot asked the participant a variety of questions in an interview-like manner. After this dialogue concluded, the robot prompted the participant to execute a task, which was used as a behavioral measure. Finally, once the participants were done with the task, they completed two self-report measures: a Robot Comfort Scale and a Robot Reaction Scale. Preliminary results show that participants had better reactions to the extroverted female robot compared to the introverted female robot. However, the opposite was true for the male robot: participants had better reactions to the introverted male robot compared to the extroverted male robot. This study demonstrates that the notion of extroverted robots yielding desirable social outcomes is not generalizable on dimensions of gender. Furthermore, humans have the tendency to feel most comfortable with extroverted female robots and with introverted male robots.