An Investigation of the Impressions of Agents Encouraging Humans to Adopt Intentional Stance by Refraining from Action

Abstract

In this study, we investigate the impression that agents refraining from action give to humans. As interactions between humans and computational agents become more common, agents interacting with humans are required to behave cooperatively. In human communication, refraining from action without consideration for others is often seen as a cooperative behavior, known as "enryo" in Japan. There are few studies examining how this behavior by agents affects human impressions. Therefore, we develop an experimental virtual environment where humans and agents engage in a cooperative task, including both a refraining agent and a non-refraining agent. We examine whether humans perceive the refraining agents as showing enryo. The results have indicated that humans tend to view the refraining agent as demonstrating enryo and have found it more anthropomorphic and likable compared with the non-(breakpoint)refraining agent.