Skip to main content
DE: diese Seite auf deutsch

New evaluation dimensions and changed weighting in the use of AI

The changes in examination practice mean that new assessment criteria must be developed for students' own work. These can be derived directly from the requirements that students have to fulfill when working properly and professionally with AI in order to successfully produce high-quality results.

Examples of assessment dimensions when using AI tools could be:

  • Assessment of AI competence through comprehensible documentation of AI use, e.g. student documentation shows that good prompts were used or adequate tools were selected. The AI output should also be critically reviewed with additional scientific sources and be revised or supplemented accordingly.
  • Uniqueness of own work, e.g. theoretical, empirical or technical content, analyses of own data, series of measurements.
  • Inclusion of personal reflection, e.g. learning log or own statements (according to Gimpel et al., 2023).

Consideration should also be given to changing the weighting of the examination criteria, e.g.:

  • Depth of argumentation, originality of solutions or reflective ability should be weighted more heavily.
  • Greater weight could be given to the oral defense of final theses, as it enables examiners to assess how thoroughly students have engaged with and understood the content.
  • Linguistic aspects, on the other hand, could be given less weight in the future if more and more students use AI to revise their work.

In conclusion, generative AI requires universities to continuously adapt their teaching content and assessment formats. By deliberately aligning examination formats with the evolving conditions of an AI-integrated world, institutions can not only enhance the validity and meaningfulness of assessment outcomes but also better prepare students for an increasingly digital professional environment. These challenges can be addressed through a combination of transparent regulations, pedagogically sound teaching and assessment design, and a culture of trust grounded in academic integrity and a forward-looking perspective.

To top