This study takes an initial step to employ learning analytics to build performance prediction models to help struggling students. It allows instructors to offer in-time intervention and support for these at-risk students. Specifically, we develop a task model to characterize the engineering design process so that the data features can be associated with the abstract engineering design phases. Teachers can integrate the feature importance ranking with the abstract task model to diagnose students’ problems for scaffolding design. The results show that the proposed approach can outperform the baseline models as well as providing actionable insights for teachers to provide personalized and timely feedback to students.
In this project, we propose a novel algorithm by incorporating the memory units in Recurrent Neural Network into a Logistic Regression (RNNLR) algorithm to identify the at-risk students and provide interpretable results in online learning settings. Variables like demographic information and different learning behaviors recorded by a learning management system were analyzed when building the predictive model. The experimental results revealed that the proposed model outperformed the baseline models for prediction accuracy and can generate interpretable results to help instructors to personalize interventions.
The present study intends to examine the links between the patterns of students’ learning activities and learning motivations based on educational data mining approaches. Various statistical analytics methods were applied to describe the process of clustering students with similar learning motivations and the associations between different educational levels and learning motivations. The experimental results show that the learning behaviors and learning styles (the procrastination levels) are significant indicators for learning motivations. Additional results from machine learning analysis showed that learning motivations can be identified by demographic information and learning activities at an early stage in the course, and there is a significant statistical association between learning motivations and final performances.