| Sumario: | Background: Self‐regulated learning (SRL) is a predictive variable in students' academic performance, especially in virtual reality (VR) environments, which lack monitoring and control. However, current research on VR encounters challenges in effective interventions of cognitive and affective regulation, and visualising the SRL processes using multimodal data. Objectives: This study aimed to analyse multimodal data to investigate the SRL processes (behaviour, cognition and affective states) and learning performance in the VR environment. Methods: This study developed a VR‐based immersive learning system that supports SRL activities, and conducted a pilot study in an English for Geography course. A total of 21 undergraduates participated. Face tracker, electroencephalography, and learning logs were used to gather data for learning behaviour, cognition and affective states in the VR environment. Results and Conclusions: First, the study identified three categories of learners (HG, MG and LG) within the VR environment who presented different behavioural engagement and SRL strategies. The HG exhibited the highest level of cognition and affective states, which resulted in superior performance in terms of vocabulary acquisition and retention. The MG, despite possessing a higher level of cognition, performed inadequately in other aspects, leading to no difference in vocabulary acquisition and retention from the LG. By collecting and mining multimodal data, this study helps to enrich the visual analysis of SRL processes. In addition, the results of this study help to dissect the problems of students' SRL in a VR learning environment. Furthermore, this study provides a theoretical basis and reference for the study of SRL development in immersive learning environments. Lay Description: What is already known about this topic?: Self‐regulated learning (SRL) is a predictive variable in students' academic performance.Students in virtual reality (VR) environments may lack monitoring and control.SRL is a complex system influenced by multiple factors.SRL in VR environments needs multimodal data to analyse the process. What this paper adds?: A VR‐based immersive system that supports SRL activities is devised.Analysis of SRL processes and performance obtained from multimodal data are discussed.Students present different behavioural engagement and SRL strategies in the VR environment.Students with different SRL behaviours present different attention, affective state, and learning performance. Implications for practise and/or policy: Effective SRL interventions should be designed within the VR environments.The mining of SRL processes with multimodal data used in the study is recommended for explaining the SRL mechanisms in VR environments.
|
|---|