Exploring the regulation of learning within mobile mixed-reality environments: Insights from dynamic engagement transitions

Analyzing transitions in multidimensional engagement states across regulated learning phases offers valuable insights into learning behaviors at a granular level. Nevertheless, the extraction of multimodal trigger signals to evaluate engagement dynamics remains a critical challenge, particularly within mixed-reality environments. Hence, this study developed a mixed-reality learning system using multi-scaffolding tools as system architectures to integrate regulatory mechanisms along the immersive learning trajectory. Specifically, the color-coded concept maps were designed as the user interface to support planning and goal orientation, while two-tier tests were incorporated as a content structuring mechanism to help learners monitor, evaluate, and adapt learning status. A comprehensive six-dimensional engagement framework was proposed to incorporate interactive, constructive, active, passive, emotional, and behavioral engagement, serving as the theoretical underpinning for the extraction of trigger signals within regulated learning phases and subprocesses. Empirical research was conducted to assess the effectiveness and learner perceptions of this novel mixed-reality learning system, utilizing a 2 × 2 mixed factorial design, with regulation type (self-regulated vs. peer-scaffolded) as the between-subjects factor and device type (smartphones vs. tablets) as the within-subjects factor. Results from 64 high school students indicated that multi-scaffolding tools significantly enhanced learning achievement across conditions. Smartphones, due to lighter weight and smaller displays, encouraged more physical behavioral engagement, while tablets, with higher resolution and larger displays, fostered greater constructive engagement. Learners in peer-scaffolded learning exhibited higher engagement transition among emotional, active, and constructive states but inefficient learning adaption, whereas learners in self-regulated learning concentrated more on task-oriented behaviors.