Anthropomorphism can be defined as attributing human-like characteristics and capabilities to non-human agents, e.g., robots. Anthropomorphism is thought to heighten acceptance of robots and to improve human-robot-interactions (HRI). No study, however, has yet yielded clear evidence for this suggestion.

The project is mainly concerned with two questions: 1) How can anthropomorphic inferences about robots be increased? 2) Does anthropomorphism actually have positive consequences for HRI?

In a first step, it is necessary to identify strategies for enhancing anthropomorphization of robots. To do so, we apply knowledge from social psychological research on human-human-interaction to HRI. For instance, social psychological research has demonstrated that we categorize people according to their age, gender or their ethnicity. Depending on whether we perceive others in terms of being in-group or out-group members, in-group bias might result. That is, for instance, people evaluate in-group members more positively and experience more positive affect toward them than toward out-group members. Most importantly, people even tend to perceive in-group members as being more human than out-group members (e.g., dehumanization). In line with these findings, we hypothesize that perceiving a robot as an in-group member vs. an out-group member would lead to more anthropomorphic inferences about the robot. To put it differently: an in-group robot would be perceived as more human than an out-group robot.

In a second step, we will explore the positive effects of anthropomorphism on HRI. We will mainly focus on three potential outcomes: increased willingness to interact with robots, reduced feelings of stress and tension in HRI, and facilitated learning about the robot.

To investigate these research questions, we apply different experimental methods to manipulate group-membership of robots, and we use diverse measures of anthropomorphism mainly derived from social psychological research on de- and infra-humanization.