Tübingen Center for Digital Education

Stephen Bodnar

My research specializes in applying advanced technologies in educational settings. It has two main aims: first, to identify aspects of instruction that matter, and second, to increase student access to high-quality practice exercises.

Currently I am involved in two projects combining E-Learning and Artificial Intelligence / Large Language Models. In the project "Using AI to topically personalise grammar learning", our team is exploring what happens when textbook-style practice materials are tailored to students' personal interests (e.g., skiers learning French might receive grammar exercises contextualised with a ski trip to the Alps). In "VOILA - Voice Operated Intelligent Learning Assistant for Gifted Children", we are developing an AI tutor that scaffolds elementary school children while they complete their homework. Because younger children tend to communicate more with speech, the tutor will be voice-interactive - this is challenging with young voices but speech technology is advancing quickly.

In my work, I enjoy looking at learning in these innovative instructional environments from a perspective that combines learning outcomes, learning behaviour, and subjective experiences, with the hope of identifying effective instructional approaches that students find meaningful and valuable.

 

April 2023 – Present
Co-head of Software Development

Tübingen Center for Digital Education (TüCeDE)
University of Tübingen, Tübingen, Germany

Jan. 2021 - Mar. 2023
Postdoctoral researcher

Department of Theoretical Computational Linguistics,
University of Tübingen, Tübingen, Germany

Mar. 2019 - Dec. 2020
Postdoctoral fellow

Department of Integrated Studies in Education (DISE), 
Faculty of Education, McGill University, Montreal, Canada

Sept. 2016 - Jan. 2018
Postdoctoral fellow

Department of Educational and Counseling Psychology (ECP), 
Faculty of Education, McGill University, Montreal, Canada

2016
PhD, Linguistics

Centre for Language and Speech Technology, Centre for Language Studies,
Radboud University, Nijmegen, The Netherlands
Thesis: Affective L2 learning experiences and ideal L2 selves in spoken CALL practice
Promotors: Professor Roeland van Hout, Dr. Helmer Strik, Dr. Catia Cucchiarini

Publications

Selected Publications

Bodnar, S. (2025). A prototype authoring tool for editing authentic texts using LLMs to increase support for contextualised L2 grammar practice. In Proceedings of the 14th Workshop on NLP
for Computer Assisted Language Learning (pp. 63-71). Tallinn, Estonia.

Bodnar, S. (2022). The instructional effectiveness of automatically generated exercises for learning French grammatical gender: preliminary results. In Proceedings of the 11th Workshop on NLP for Computer Assisted Language Learning (pp. 10-22). Louvain-la-Neuve, Belgium.

Bodnar, S., Cucchiarini, C., Penning de Vries, B.W.F., Strik, H., & Hout, R.W.N.M. van. (2017). Learner affect in computerised L2 oral grammar practice with corrective feedback. Computer Assisted Language Learning, 30(3-4), 223-246. dx.doi.org/10.1080/09588221.2017.1302964

Bodnar, S., Cucchiarini, C., Strik, H., & van Hout, R.W.N.M. van (2016). Evaluating the motivational impact of CALL systems: current practices and future directions. Computer Assisted Language Learning, 29(1), 186-212. dx.doi.org/10.1080/09588221.2014.927365

Bodnar, S., Cucchiarini, C. & Strik, H. (2011). Computer-assisted grammar practice for oral communication. In Proceedings of the 3rd International Conference on Computer Supported Education (CSEDU) (pp. 355-361). Noordwijkerhout, The Netherlands.

Research Grants

Main applicant: 2023-2026, with Katharina Wendebourg and Cora Parrisius.
Agency: LEAD Intramural Research Grant 
Title: Using AI to topically personalise grammar learning
Amount: €9297

Co-applicant: 2022-2023, with Elizabeth Bear (main applicant) and Xiaobin Chen.
Agency: LEAD Intramural Research Grant
Title: Can state-of-the-art commercial ASRs be used for ICALL systems with spoken language?Amount: €9100

Principal Investigator: 2019-2020
Agency: Fonds de recherche du Québec – Société et culture
Title: A pluggable-content approach to computerized French grammar instruction (postdoctoral grant)
Amount: $90,000