Investigation of the effects of a robot's proactive involvement on users' evaluation of created artifacts in human–robot joint tasks

In interactions with dialogue robots, positive feedback, such as praise or agreement, has often been used to shape users' impressions of the robot. However, when a robot is involved in a user's task process, merely affirming the user's choices may be insufficient to influence how users evaluate the outcomes of joint tasks. This study focused on a robot that engages in proactive involvement during a task instead of only providing positive feedback. We investigated how proactive involvement by a robot affected user evaluations in a joint creation context. Using an open-ended character-creation task with no single correct answer, we compared a robot that engaged in proactive involvement with one that responded only with positive feedback. The results showed that a proactive robot was evaluated to have higher anthropomorphism, animacy, and perceived psychological closeness (Inclusion of Other in the Self (IOS) Scale) than a robot that provided positive feedback alone. In addition, proactive involvement was associated with behavioral preference for the jointly created artifact. In conclusion, these findings suggest that proactive involvement, as a form of task-level involvement by robots, can influence both user perceptions of the robot and their evaluations of artifacts created through joint tasks.