Evaluating and Improving Pedagogical Fit in LLM-Based AI Tutors with the Pedagogical Suitability Index

The authors argue that existing evaluations of AI tutors overemphasize answer correctness while ignoring instructional fit. They propose the Pedagogical Suitability Index as a metric to capture alignment with learner foundation, curriculum progression, and timing of concept introduction. The paper appears to present both an evaluation framework and methods for improving LLM tutor responses using this index. This work is relevant to the growing deployment of LLM-based tutoring systems in real classrooms, where context-aware help is critical.