Assessment of Student Engagement in Undergraduate Blended Learning: A Case Study on the Database Principles and Applications Curriculum

Assessing student engagement in blended learning is critical for enhancing instructional effectiveness. While existing research has predominantly focused on face-to-face or online learning environments in China, studies evaluating engagement in blended learning curricula in higher education remain limited. This study investigates student engagement across three key dimensions—behavioral, cognitive, and social interaction—using the "Database Principles and Applications" course as a case study. An engagement assessment model was developed by analyzing students' learning processes in a blended environment, identifying 11 key indicators, and incorporating machine learning techniques. A key contribution of this study is the exploration of teachers' ongoing pedagogical improvements as a significant factor influencing engagement assessment. By comparing two evaluation models, batch and incremental, the study seeks to identify the most effective assessment approach. Experimental results demonstrate that the incremental learning Random Forest algorithm improves prediction accuracy, addresses stability issues when new data is introduced, and supports the monitoring of student learning behaviors across multiple semesters.