Effects of Implicit Role Differentiation on Performance in Dyadic Joint Tasks

Abstract

Joint action often requires individuals to coordinate without explicit communication, relying instead on mutual adaptation. While role differentiation has been considered a key factor in successful coordination, its relationship with performance remains unclear and context-dependent. In this study, we investigated how different forms of role differentiation contribute to dyadic coordination using a joint force-production task with familiar pairs. Participants were required to collaboratively adjust their force output to match a intermittently changing target based on shared visual feedback reflecting only the average force of the pair. This setup prevented access to individual contributions and imposed conditions without verbal communication, explicit rules, or top-down direction. We quantified role differentiation in two domains: (1) Overall force magnitude and (2) Responsiveness to intermittent target changes. The results showed that clear and stable role differentiation emerged in overall force magnitude and was aligned with participants’ subjective reports. However, this form of differentiation was not associated with task performance. In contrast, role differentiation in responsiveness to intermittent changes was less pronounced and not consciously perceived, yet it was associated with improved performance. Furthermore, this differentiation exceeded chance levels, suggesting the presence of implicit coordination processes. These findings indicate that not all role differentiation equally contributes to successful coordination; rather, subtle and unperceived differentiation in adaptive responses may play a critical role. Our results highlight the importance of capturing implicit coordination dynamics in collaborative interactive systems. Understanding how such role differentiation emerges may inform the design of intelligent systems that support flexible, real-time human–human collaboration and may provide insights into adaptive role allocation based on individual coordination characteristics in future human–agent teams.