Q-matrices are a cornerstone of theory-driven assessment and learning analytics, making item demands and students' underlying knowledge components and misconceptions explicit and actionable. However, Q-matrices are typically crafted by experts, making them time-consuming to build, prone to subjectivity, and difficult to validate empirically. We propose a framework for human-AI Q-matrix refinement in which large language models (LLMs) generate candidate Q-matrices using structured, misconception-aware prompting, and NeuralCDM provides an empirical evaluation layer to compare candidates based on how well they explain student response data. We apply the framework to a thermodynamics assessment dataset and benchmark locally deployed LLMs against cloud-served models. Results show that iteratively refined LLM-generated Q-matrices can exceed expert-baseline model fit (AUC 0.780 vs. 0.717), and that locally deployed models achieve comparable performance to cloud APIs, supporting privacy-preserving deployment.
@article{arxiv.2604.16398,
title = {A Framework for Human-AI Q-Matrix Refinement: A NeuralCDM Evaluation},
author = {Ying Zhang and Ningxi Cheng and Yizhu Gao and Hongmei Li and Lehong Shi and Nicholas Young and Geng Yuan and Xiaoming Zhai},
journal= {arXiv preprint arXiv:2604.16398},
year = {2026}
}