Multimodal Learning Analytics: An Overview of the Data Collection Methodology

Since, multimodal learning analytics has shared a common interdisciplinary approach with learning analytics utilizing technologies like low-cost sensors and wearable devices, artificial intelligence (AI), and machine learning (ML) to provide real-time feedback to students and instructors. MMLA focuses on research within intermediary spaces— between research and classroom application, bridging private and public learning contexts by tracking students’ interactions with social media and connected devices, and exploring context-aware and adaptive technologies. However, there is a dearth of evidence demonstrating the successful application of multimodal learning findings in classroom settings on a larger scale. The encroachment on students’ privacy due to the intrusive nature of multimodal data collection methods, often involving sensors and recording technologies, remains a substantial concern that necessitates attention. This study systematically reviews multimodal learning analytics, examining aspects such as data collection rationale, data scope, study contexts, ethical considerations, commercial interests, and downstream data uses. Findings emphasize the need for greater ethical scrutiny, reevaluation of data collection rationale, and comprehensive privacy protections.