Rubric-Based Explainable Feedback with AI-Assistance

Timely formative feedback plays a crucial role in fostering self-regulated learning and promoting deep learning strategies. Research highlights that feedback enhances student engagement, satisfaction, and learning outcomes, particularly in generative tasks like ePortfolios and other multimodal compositions. However, providing detailed formative feedback is resource-intensive for teaching staff. Our presentation outlines a general framework for multimodal formative assessment analysis. The approach combines task-specific assessment rubrics and metrics for explainable and formative feedback, which in turn are used to create an assessment as well as explanations and suggestions. This presentation will also discuss the steps needed to evaluate and refine this process for real-world classroom applications.

Literatur: Darling-Hammond, Kia; Darling-Hammond, Linda; Byard, Eliza (2022): The civil rights road to deeper learning. five essentials for equity. New York: Teachers College Press. Maya, Fatima; Wolf, Karsten D. (2024): An Architecture for Formative Assessment Analytics of Multimodal Artefacts in ePortfolios Supported by Artificial Intelligence. In: Muhittin Sahin und Dirk Ifenthaler (Hg.): Assessment Analytics in Education. Designs, Methods and Solutions. 1st ed. 2024. Cham: Springer International Publishing; Imprint Springer (Advances in Analytics for Learning and Teaching), S. 293–312. Thurlings, Marieke; Vermeulen, Marjan; Bastiaens, Theo; Stijnen, Sjef (2013): Understanding feedback. A learning theory perspective. In: Educational Research Review (9), S. 1–15. Wolf, Karsten D.; Maya, Fatima; Heilmann, Lisanne (2024): Explainable Feedback for Learning Based on Rubric-Based Multimodal Assessment Analytics with AI. In: Natalie Kiesler und Sandra Schulz (Hg.): Workshopband der 22. Fachtagung Bildungstechnologien (DELFI). 9.-11. September 2024; Fulda, Deutschland. Fulda: Gesellschaft für Informatik e.V, S. 283–292.