Learning Analytics
Learning analytics — the measurement, collection, analysis, and reporting of data about learners and their contexts — is the integrating thread across all my research. From the earliest work on predicting underachieving medical students through their online activity traces, I have been interested in what educational data can reliably tell us about learning, and what it cannot.
A persistent theme is the tension between prediction and explanation. My research has shown that early prediction of at-risk students is feasible using engagement indicators from learning management systems, but that the accuracy and the predictors themselves vary substantially across courses and individuals. Through single-paper meta-analyses pooling results across many courses, I have identified which variables capture student performance reliably and consistently, and where individual differences undermine cross-context generalization. This has led to a deeper engagement with person-centered methods — latent class analysis, latent profile analysis, clustering, and mixture models — that respect heterogeneity rather than averaging it away. I have also contributed methodological tutorials on structural equation modeling, epistemic network analysis, and the broader learning analytics pipeline, aiming to make these techniques accessible and reproducible for educational researchers worldwide.
Selected Publications
- How Learning Analytics Can Early Predict Under-Achieving Students in a Blended Medical Education Course (2017)
- When, How and For Whom Changes in Engagement Happen: A Transition Analysis of Instructional Variables (2023)
- Modelling Within-Person Idiographic Variance Could Help Explain and Individualize Learning (2023)
- A Modern Approach to Transition Analysis and Process Mining with Markov Models in Education (2024)
- Structural Equation Modeling with R for Education Scientists (2024)
- A Literature Review of Empirical Research on Learning Analytics in Medical Education (2018)
- Automating Individualized Machine Learning and AI Prediction Using AutoML (2026)