Exploring Human-Conversational Agent Team Dynamics

Technological advancements empower conversational agents (CAs) to perform team tasks that, until recently, were exclusively handled by humans. Thus, CAs contribute to team collaborations. Yet it remains unclear whether humans perceive CAs as part of the team, and whether human-CA collaborations are comparable to human-only teams. This work investigates the impact of a chatbot adopting a Scrum Master assistant (SM) role on emerging team dynamics in a lab experiment with 150 participants. The preliminary findings suggest that CAs are attributed lower team membership than humans. Ingroup perception revealed to be a crucial antecedent for individual satisfaction with the SM, resulting in higher team mental model formation and trust in the SM. The findings suggest that ingroup perception explains the difference of team dynamics across various teams (i.e., human-only and human-CA teams). With additional data, we will contribute to research on ingroup identification in human-CA team dynamics.