Skip to main content Skip to main navigation


Performance by Preferences–An Experiment in Language Learning to Argue for Personalization

Leo Sylvio Rüdian; Niels Pinkwart
In: Artificial Intelligence in Education. International Conference on Artificial Intelligence in Education (AIED-2023), July 3-7, Tokyo, Japan, Pages 347-352, ISBN 978-3-031-36336-8, Springer Nature Switzerland, 6/2023.


Personalizing online courses is of high interest due to heterogeneous learners. Mainly, personalization is limited to learning content, or difficulty levels. Often, courses are just optimized resulting in one version, the one-size-fits-all variant. However, learners are considered as one cohort, independently of sub-groups, and their existence is seldom further analyzed. In this paper, a metric is introduced, the so-called preference discrimination index, as an indicator for a potential bias caused by the selection of instructional methods. We examine two versions of a 45min language learning online course, which cover the same learning content, but one version differs in instructional methods, enriched by simulations. The result shows that the “traditional” course without simulations discriminates more for considered preferences, indicating the need for improvements.