협력학습에서의 자기조절학습 측정을 위한 멀티모달 학습분석(MMLA) 활용: 체계적 문헌고찰

초록

Multimodal learning analytics (MMLA) has emerged as a promising approach for capturing self-regulated learning (SRL) in collaborative contexts. However, the systematic understanding of how different data modalities and analysis techniques reflect regulation processes remains limited. This systematic review analyzed 21 peer-reviewed empirical studies with four objectives: (1) identifying the research purposes and contexts in which MMLA has been applied to measure SRL in collaborative learning, (2) examining the theoretical foundations underpinning these studies, (3) analyzing the types of multimodal data collected and analytical methods used, and (4) analyzing the methods of multimodal data integration. The findings showed that most studies focused on diagnosing and modeling SRL in synchronous, face-to-face, collaborative settings. The theoretical grounding was dominated by Hadwin et al.’s (2011) framework of socially shared regulation. Contextual, behavioral and physiological data were triangulated across studies using advanced analytic techniques such as time-series analysis, process mining, and deep learning-based prediction models. This review outlines current practices and highlights MMLA’s potential to provide granular, theory-aligned insights into SRL in collaborative learning.

키워드