A Systematic Review of Multimodal Learning Analytics Design Models and Frameworks

Multimodal Learning Analytics (MMLA) systems hold immense potential for understanding and shaping learning experiences. However, the lack of standardized design models hinders the consistent and effective development of these systems. This systematic review addresses this gap by identifying and analyzing existing MMLA design models and frameworks. We employed a rigorous search strategy aligned with established guidelines to identify relevant studies published in the past decade. Following a qualitative approach, the review combined narrative synthesis and thematic analysis to extract and synthesize key findings. Our analysis revealed a diverse landscape of MMLA design models and frameworks, varying in their scope (specific learning activities vs. comprehensive MMLA system design), level of detail (high-level guidance vs. specific steps), and development process (theoretical foundations vs. practical experience). Notably, several models addressed key design considerations and core commitments emphasized by recent research (e.g., data privacy, learner agency, inclusive learning environments). More importantly, the aggregation of these identified models suggests promise for the development of a more comprehensive design model. This is because individual models cover distinct areas and aspects with some intersections. The review also identified recurring themes related to factors influencing MMLA system design, including usability, scalability, and ethical considerations. Finally, we provide a discussion on potential strategies for a concrete development, offering valuable insights for researchers, developers, and educators seeking to harness the power of MMLA to improve learning outcomes.