From Interactional Shifts to Higher-Order Consensus: An Exploratory Analysis of Collaborative Learning Using an Event-Locked Time-Window Approach

Research on collaborative learning has commonly conceptualized consensus as a learning outcome or an overall indicator of collaboration quality, with limited attention to when consensus emerges during interaction and how it is formed. To address this gap, the present study proposes an exploratory, event-locked time-window perspective that conceptualizes higher-order consensus as an observable interactional shift within collaborative discourse, rather than as a final state reached at task completion.Analyses of two authentic small-group speaking tasks indicate that, in prompt-triggered cases, higher-order consensus does not accumulate gradually through extended discussion. Instead, it emerges in close temporal proximity to a clearly identifiable interactional event and is concentrated within a very short interval following that event (within one to two turns after t₀). In contrast, in naturally generated cases, higher-order consensus can also be achieved, but its formation follows a more gradual trajectory and lacks a clearly identifiable turning point.Further analyses suggest that the distinction between these two pathways lies not in whether consensus is ultimately achieved, but in whether the interactional shift leading to consensus manifests as a temporally concentrated change. Without advancing predictive models or causal claims, this study employs behavior-level, event-locked analysis to characterize the temporal features of higher-order consensus as an interactional turning point, offering a clear and operational process-oriented perspective for examining consensus formation in collaborative learning.