ScholarWorks@숙명여자대학교

초록

This study examined how learners’ regulation behaviors are identified, interpreted, and supported through data-informed approaches in Multimodal Learning Analytics(MMLA). To this end, the study analyzed research on Self-Regulated Learning(SRL) and Socially Shared Regulation of Learning(SSRL) based on the MMLA workflow framework. Twenty-five studies published between 2022 and 2024 were selected from the SCOPUS database using keywords such as ‘multimodal’, ‘learning analytics’, and ‘regulation’. The results indicate that SRL research primarily utilized physiological and behavioral data to explore individual cognitive and emotional regulation processes in a time-series manner. In contrast, SSRL research tended to analyze the structure of group shared regulation by integrating social interaction data, including speech, behavior, and AI logs. The discussion highlights the integrative trends between SRL and SSRL, the variation in data types depending on learning contexts (e.g., face-to-face vs. online), the emergence of data fusion techniques, the application of diverse analytical methods, and the connection of data utilization to practical learning support and instructional design. Although a fully integrated explanatory framework for SRL and SSRL has not yet been established, this study confirms that recent research is increasingly fusing multimodal data to analyze learners’ regulation processes with greater precision and utilizing the findings for real-time feedback and instructional design support.

키워드