Rapport has been argued to be one of the central, if not the central, constructs necessary to understanding successful helping relationships and to explaining the development of personal relationships. Through systematic experimentation with humans and virtual (computer-based) humans, this project seeks to deepen understanding of the factors underlying rapport by looking more closely at the synchronization (i.e., process entrainment). The research will employ emotionally expressive virtual humans that are responsive to human participants' (i) respiratory patterns, (ii) facial feature and gaze tracking, and (iii) optical flow. If "unilateral entrainment" (of humans to virtual humans) produces rapport-like social consequences, it would greatly simplify interface agent design, obviating the need for sophisticated sensing and generation of matching behavior. Alternatively, it may promote certain undesirable social attributions -- incompetence, resentment, and dislike -- that could conflict with the goals of the application. This project investigates this fundamental question for human-computer interaction and virtual agent design.
This technology has the potential to facilitate the establishment of rapport between humans and virtual humans and to achieve the broader range of socially desirable consequences including improved computer-mediated learning and health interventions. As rapport is a key social moderator found to influence outcomes in a variety of settings that require teaching or persuasion, this research could inform the development of virtual training in the areas of negotiation, conflict resolution and cultural awareness. Finally, this project has broad dissemination potential given that rapport is inherently interdisciplinary, involving the domains of social psychology, business, communication and computer science.
