Use of Praise and Punishment in Human-Robot Collaborative Teams

Robots, specifically androids, become increasingly important in the consumer market where they are marketed as toys or companions, as well as in the industry, where they will increasingly often play the role of a co-worker. The developers in various robotics communities are divided about design issues in these companion-worker androids. While some robot developers believe people will work more effectively with humanoid robots in the role of companion or co-worker because of a more natural interaction, others think it's necessary to maintain a machine-like interface to avoid distractions. Consequently, the ability of humans to coordinate and interact with robots, and human perceptions and actions based on varying levels of humanlike robot interfaces are of great interest. This paper presents preliminary results from a study that investigated how people use praise and punishment in a collaborative game scenario. Subjects played a game together with humans, computers, and anthropomorphic and zoomorphic robots. They could give plus points and minus points as praise and punishment for correct or wrong partner answers. Results show that praise and punishment were used the same way for computer and human partners. Yet robots, which are essentially computers with an embodiment, were treated differently. Very machinelike robots were treated just like the computer and the human; robots very high on anthropomorphism/zoomorphism were praised more and punished less