Seyma Nur Ertekin

Faculty of Social and Behavioural Sciences
Psychology Research Institute
Psychological Methods
University of Amsterdam


Cognitive Modeling Meets Educational Data Science

Implications from cognitive science have significantly contributed to the understanding and improvement of education. One important contribution is the cognitive models and their applications to the domain of learning and memory. This project aims to combine theories of (mathematical) cognition with Bayesian modeling and large-scale educational data from the adaptive online platform, Prowise Learn. By incorporating individual differences and focusing on motivation, quitting behavior, and the domain of forgetting, we will enhance the statistical modeling approach.

Prowise Learn is an adaptive learning platform that currently employs overall performance to select the next learning items. By applying cognitive modeling techniques and using cognitive architectures, adaptive item selection can be significantly improved. By fitting cognitive models to each learner’s data, we can monitor specific cognitive parameters related to memory, activation, and motivation. This approach opens up a new way of adaptive testing where adaptivity is not only steered by ability but also by cognitive and motivational processes so that each learner is challenged based on their cognitive profile. Furthermore, this fine-grained tracking can be used for an optimal formative assessment, including theory-driven learning analytics for teachers. The results will contribute to a deeper understanding of individual differences, motivation, and forgetting in the context of learning. The project’s findings will empower educators to make data-informed instructional decisions, fostering enhanced learning experiences for all learners.

Prof. Dr. H. L. J. van der Maas
Dr. D. Matzke
Dr. A.D. Hofman
Dr. J.M. Haaf

Financed by
University of Amsterdam

December 2022 – December 2026