Effect of Different Listening Behaviors of Social Robots on Perceived Trust in Human-robot Interactions

摘要

Abstract With the increased use of social robots in prominence and beyond functional performance, they are expected to foster trust and confidence in people. Various factors involve providing social robots with more trustworthy behavior. This study investigated whether the listening behavior of a social robot can affect the perception of being trustworthy in human–robot interaction. Therefore, we designed four different listening behaviors, including nonactive listening, active listening, active empathic listening, and verbal-only empathic listening, for a social robot and evaluated the impact of each behavior on the participants’ likelihood of trusting the robot, using a between-subject design. Participants in the four conditions conversed with a robot that simulated one of the listening behaviors, and their general, cognitive and affective trust toward the robot was measured. The results indicated that active empathic listening behavior provided the participants with the highest impression of trustworthiness, specifically in affective trust. Both active listening and active empathic listening were evaluated higher than nonactive listening in general, affective, and cognitive trust. However, active empathic listening behavior was differentiated from active listening behavior only in terms of affective trust. For verbal and nonverbal dimensions of listening behaviors, it was confirmed that nonverbal behaviors such as nodding, body movement, and eye gaze along with verbal behaviors, had a significant effect in eliciting higher affective trust in human-robot interaction. Consequently, we concluded that designing social robots with active (empathic) listening behavior can enhance trust perception in human-robot interaction in different fields such as education, healthcare, and business.