Improving Interactions with Healthcare Robots

Abstract

Background and Purpose: Social robots are currently being developed for use in healthcare environments. To date, little research has investigated how healthcare robots should behave in order to promote positive patient outcomes. Research on physician behaviour may inform the design of healthcare robots within this context. The central aim of this thesis was to investigate the effects of key healthcare robot communication behaviours on user perceptions and behaviours in simulated healthcare interactions. Methods: A literature review on the effect of healthcare professionals’ communication behaviour was undertaken, including a review of these behaviours within human-robot interaction (HRI) research. This was followed by four HRI studies. Study one examined the effect of robot smiling and first name use using a within-subjects design. Utilising a between-subjects design, study two investigated the effect of robot self-disclosure, voice pitch, and forward body lean, while study three examined the effects of robot humour. Study four investigated robot empathy statements and head nodding, using a between-subjects online video design. Results: The review found several key physician communication behaviours for study within HRI research; namely humour, self-disclosure, facial expressions, empathy statements, and body posture. The experimental research found robot smiling to significantly improve user attitudes, smiling behaviours and perceptions of robot friendliness and mind. Both robot smiling and first name use significantly increased user perceptions of robot personality. Robot forward lean and self-disclosure significantly increased user eye contact behaviours and robot self-disclosure significantly increased user laughter. Conversely, robot voice pitch significantly increased user ratings of the robot as boring and unstimulating. Robot humour significantly increased participant laughter and ratings of robot likeability, safety, empathy, and personality. Robot verbal empathy significantly increased participant perceptions of satisfaction, robot trust, and robot empathy, while significantly decreasing robot distrust. Finally, robot head-nodding had no significant effect on outcome variables. Conclusions: Similar to human healthcare professionals, healthcare robots should use specific communication behaviours to improve user outcomes. These include smiling, forward lean, self-disclosure, humour, and empathy statements. Future researchers should consider the re-examination of these and other communication behaviours, using mixed-age samples, patient populations, and within the context of natural medical environments.