Skip to content
Review Open access

A Review of Multimodal Data-Driven Evaluation Research on Blended Teaching Models

2026 · Advances in Educational Technology and Psychology · 0 citations · 10 references

Abstract

: Blended teaching has become the new normal in higher education, yet its evaluation remains challenging due to the distribution of learning behaviors across online and offline spaces. Multimodal learning analytics offers a new paradigm for precise evaluation. This paper systematically reviews research on multimodal data-driven evaluation of blended teaching, examining theoretical foundations, data sources, methodologies, applications, and challenges. The review reveals that multimodal evaluation has evolved from single-dimensional behavioral measurement toward comprehensive assessment encompassing cognition, emotion, behavior, and social interaction. Methods have shifted from traditional questionnaires to machine learning and multimodal fusion, while evaluation content has moved from outcome-oriented to process-oriented assessment. Knowledge graphs are emerging as tools for knowledge visualization and cognitive diagnosis within multimodal frameworks. However, technical bottlenecks in data fusion, insufficient model interpretability, and ethical concerns remain critical challenges. Future research should focus on developing interpretable models, advancing educational multimodal fusion algorithms, and establishing frameworks that balance effectiveness with ethical considerations.

Read PDF

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.