A Hierarchical Model for Human-Robot Interaction

This paper proposes a hierarchical model for human interaction with social robots which consists of three tiers, each tier based on an existing theory of human-robot interaction. The three theories which comprise the model are robot anthropomorphism which, among others, is based on the physical appearance of the robot; Social Identity Theory, which discusses how a robot may be classified into social categories; and at the lowest tier of the model, the Computers as Social Actors Theory, in which social rules mediate human-robot interaction. The top level of the model is based on the initial impression that individuals form of robots which is determined primarily by the robot’s physical appearance. At the mid-tier of the model, Social Identity Theory is used to describe how people place robots into social categories, which among others can be based on the perceived race, ethnicity, and gender of the robot. An important consequence of categorizing robots by a social class is that the category may determine whether the robot is judged to be an in-group or out-group member which can lead to different behavioral interactions between humans and robots as described in the third tier of the model. According to the hierarchical model, communication with a robot occurs at the lowest tier of the model and is guided by the process of robot anthropomorphism, Social Identity Theory, and by predictions from the Computers as Social Actor paradigm.