Hannes Seifert has been awarded a "Fellowship for Innovation in Digital Higher Education"

Since the 2024/2025 winter semester, the University of Erfurt has been one of only two institutions in Germany to offer a dual degree programme for trainee teachers. From the third semester onwards, students spend two days a week at a partner school and three days at the university, in order to closely integrate theory and practice and to mitigate the “culture shock” they may experience when starting their careers.

A video-based learning format, developed in collaboration with the Erfurt School of Education at the University of Erfurt and the relevant subject-specific teaching methodology departments, is designed to support students in reflecting on their own teaching experiences from a subject-specific pedagogical perspective: they record their own lessons and then analyse selected sequences together in their courses. “This format is seen as beneficial, but it comes up against a key problem,” explains Hannes Seifert. “For data protection reasons, videos featuring real school classes may only be viewed within the respective course by a small group of students and must be deleted afterwards. Sustainable, cross-semester use is therefore not possible, and valuable teaching material is lost. This is where I intend to focus my efforts: over the next two semesters, I aim to build up a permanently accessible, data-protection-compliant collection of teaching sequences from Mathematics lessons, in order to systematically enhance the professional competence of students on the dual teacher training programme for mainstream schools.”

To this end, suitable short sequences with high pedagogical potential are to be identified from existing teaching videos, abstracted into prompts and anonymised. Using AI video generators, new video vignettes featuring simulated teachers and learning groups will then be reconstructed from these and gradually compiled into a collection of video vignettes that can be used across semesters in the courses. The video vignettes on various topics in Mathematics (e.g. negative numbers or systems of equations) will each be linked to appropriate analysis tasks (e.g. relating to cognitive processes or misconceptions).

“We are deliberately including open-ended or ambiguous teaching situations, rather than just best-practice examples, in order to stimulate discussions with students about possible alternative courses of action,” says Hannes Seifert. “The AI-generated vignettes are not intended to replace real video reflection, but to complement it effectively and thus ensure the long-term availability of material that was previously lost for data protection reasons.”