When learning analytics dashboard is explainable: An exploratory study on the effect of GenAI-supported learning analytics dashboard

Explainability matters more for learning than for task performance: While students using the explainable AI dashboard didn't produce better abstracts than those with basic visual feedback, they scored significantly higher on knowledge tests about abstract writing principles (p=.026), suggesting that understanding why AI makes suggestions is essential for genuine learning, not just completing assignments.

AI transparency supports self-regulated learning: The study found that explainable AI features—designed around Self-Regulated Learning theory with visual, explainable, and interactive layers—helped students develop deeper conceptual understanding and transferable skills they could apply beyond the immediate task, rather than just following AI suggestions mechanically.

Simple AI feedback is insufficient for educational contexts: The research demonstrates that providing students with AI-generated feedback alone is not enough; without explanations of the reasoning behind that feedback, students may complete tasks successfully but miss opportunities to develop critical thinking and independent academic skills.

Responsible AI design requires theoretical grounding: The study's success in improving conceptual understanding came from deliberately designing the AI system around established educational theory (Self-Regulated Learning) and XAI principles, rather than simply adding AI to existing tools—suggesting that responsible AI in education requires intentional pedagogical design.