Machine Learning-Based Feedback Support for Blended Learning in Medical Studies

Blended Learning is used to prepare medical students for real-life situations, and for effective learning, students and teachers need detailed feedback.

Abstract

Virtual Patients (VPs) have been used with great success for blended learning courses in the education of medical students. While VPs are perceived by users very positively overall, we intend to improve the way feedback is given to both students and their teachers. For this purpose, two prototypes will be developed in this project to support effective feedback within a blended learning course using VPs, and they will be pilot tested by the respective stakeholders, both teachers and students.

The first prototype will generate and visualize an automated individual assessment of a student’s narrative summaries on the VP tasks by means of text analysis (Natural Language Processing NLP). The second prototype will visualize different learning analytics data of the respective students, i.e., all interactions with the VP including all provided answers to questions.

Both students and teachers will be evaluated in focus group studies how helpful they perceive this new form of feedback (via the two prototypes) from their respective points of view when integrated into a blended learning course for medical students.

(Zwischen-) Ergebnisse und Infos zum Projektstand

The project starts in January 2025. As a first step, data from students› VP encounters will be collected and analyzed.

Translation

The developed prototypes will be implemented and evaluated by means of a pilot project within a regular blended learning course for medical students at the Medical Faculty of the University of Bern. If evaluated positively, it is intended to establish both developed prototypes as standard in these courses.

Gesamtleitung

Projektmitarbeit

Beteiligte Institutionen