Impact of Robot Appearance and Attitude Changes on Human Impressions: a Preliminary Study

Abstract

In this paper, we present an interaction model aimed at creating a friendly impression through the adjustment of robot attitudes based on their appearances. Establishing a friendly rapport is crucial for communication robots to effectively engage and heal people. In interpersonal communication, the gain-loss theory highlights a phenomenon called the gain-loss effect. This effect suggests that people find others more appealing when their attitude changes rather than when it remains the same. To investigate this phenomenon in human-robots interaction, we conducted an experiment where robots with diverse appearances guided users while adapting their attitudes to match their language and actions. The results suggest the occurrence of the gain-loss effect, as the impression of the robots’ behavior changed alongside their attitude adjustments. Specifically, robots exhibiting a gain condition had a positive influence on impressions and conveyed a sense of human-likeness. In contrast, robots in the loss condition had no impact on positive impressions, and conveyed less human-likeness.