Machine learning for the educational sciences

Abstract

Machine learning (ML) provides a powerful framework for the analysis of high‐dimensional datasets by modelling complex relationships, often encountered in modern data with many variables, cases and potentially non‐linear effects. The impact of ML methods on research and practical applications in the educational sciences is still limited, but continuously grows, as larger and more complex datasets become available through massive open online courses (MOOCs) and large‐scale investigations. The educational sciences are at a crucial pivot point, because of the anticipated impact ML methods hold for the field. To provide educational researchers with an elaborate introduction to the topic, we provide an instructional summary of the opportunities and challenges of ML for the educational sciences, show how a look at related disciplines can help learning from their experiences, and argue for a philosophical shift in model evaluation. We demonstrate how the overall quality of data analysis in educational research can benefit from these methods and show how ML can play a decisive role in the validation of empirical models. Specifically, we (1) provide an overview of the types of data suitable for ML and (2) give practical advice for the application of ML methods. In each section, we provide analytical examples and reproducible R code. Also, we provide an extensive Appendix on ML‐based applications for education. This instructional summary will help educational scientists and practitioners to prepare for the promises and threats that come with the shift towards digitisation and large‐scale assessment in education.

Closing the loop – The human role in artificial intelligence for education

“…For instance, ML and AI could be used to predict dropout rates in college or learning success for a course, but the underlying mechanisms might remain hidden from us. Nevertheless, the recent trend toward interpretable ML addresses the criticism of conventional ML of merely providing predictions and emphasizes transparency of the inner workings of ML models to better understand ML-guided decision-making (for a deeper methodological discussion, see Hilbert et al, 2021 ). This is especially relevant when studying learning processes, as it is crucial to find out which individual variables or aspects of an intervention positively or negatively influence learning success.…”

Misserfolge beim digitalen Lernen verhindern: „Predictive learning analytics“ am Beispiel einer Web-App für Grammatik und Rechtschreibung

“…MOOCs (massive open online courses), frei zugängliche Online-Kurse von großen, weltweit agierenden (Online-)Universitäten zu Themen überwiegend aus dem MINT-Bereich (Hellas et al 2018;Hu 2022). Als Analysetechniken werden vor allem Verfahren des maschinellen Lernens eingesetzt (Hilbert et al 2021). Beispielsweise eignen sich machine learning classifiers, um komplexe, non-lineare Zusammenhänge zwischen Merkmalen und einer Zielvariable zu modellieren.…”

Using Machine Learning to Predict Mathematical Performance

“…In some fields (e.g., information and health sciences, cf. Hilbert et al, 2021), application of ML already constitutes the standard, whereas in educational sciences, there has not yet been such a coming shift in analytical paradigms. Although classical statistical regression methods, as mentioned, can only represent (generalized) linear relationships between a restricted number of variables, ML methods typically include more high-dimensional relationships without limiting the number of variables.…”

“…Further advantages of ML are the easy detection and representation of non-linear relationships (such as depicted in Figure 2). In addition, the incorporation of elaborate (nested) resampling techniques in combination with prediction-centered model evaluation, predestines ML approaches for the development of thoroughly validated models that are replicable with novel data (Hilbert et al, 2021).…”

“…In general, educational research is particularly engaged in identifying and evaluating important predictors of student achievement. Because the success of educational systems is of utmost relevance for societies (e.g., Hanushek & Woessmann, 2012), a flurry of empirical research consequently investigates achievement measures such as mathematical, scientific, or reading competencies as dependent variables (Hilbert et al, 2021).…”