Relationship Building in Human-Machine Interaction : (A Proof of Concept on the Influence of Self-Disclosure on Rapport in Chatbot Coaching using Psychophysiological Measurements)

This study investigates the influence of the disclosure behaviour of a coaching chatbot on the development of rapport in human-machine interaction. Accordingly, we create a proof of concept for the use of psychophysiological measurement methods to investigate rapport in human-machine interaction. Regarding the psychophysiological measurements, we develop a stimulus in the form of a video recording of a simulated chatbot coaching session on the topic of exam anxiety. Using a between-subject design we form two test groups of ten people each, one group watching a video with a self-disclosure chatbot and the other watching a video with a chatbot showing information-disclosure behaviour. In addition to recording the body’s stimulus reactions, we use a questionnaire to subjectively record rapport. Rapport – originally a concept from psychology – describes a relationship of trust between people or people and machines. The analysis results of the psychophysiological measurement indicate a stronger rapport with the self-disclosure chatbot. However, this effect is not reflected in the questionnaire results. The proof of concept, thus, demonstrates the fundamental feasibility and effectiveness of psychophysiological measurements. The results make a contribution to research in human-machine interaction using the example of chatbot coaching and provide indications of relationship-building factors.