Measuring Competency, Not Performance: Item-Aware Evaluation Across Medical Benchmarks
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866911568113958912 |
|---|---|
| author | Luo, Zhimeng Wu, Lixin Frisch, Adam He, Daqing |
| author_facet | Luo, Zhimeng Wu, Lixin Frisch, Adam He, Daqing |
| contents | Accuracy-based evaluation of Large Language Models (LLMs) measures benchmark-specific performance rather than underlying medical competency: it treats all questions as equally informative, conflates model ability with item characteristics, and thereby produces rankings that vary with benchmark choice. To address this, we introduce MedIRT, a psychometric evaluation framework grounded in Item Response Theory (IRT) that (1) jointly models latent competency and item-level difficulty and discrimination, and (2) includes benchmark integrity validation to ensure items within each topic measure a single, coherent underlying ability. We prospectively evaluate 71 diverse LLMs on a USMLE-aligned benchmark across 11 medical topics. As internal validation, MedIRT correctly predicts held-out LLM responses on unseen questions with 83.3% accuracy. As external validation, IRT-based rankings outperform accuracy-based rankings across 6 independent external medical benchmarks -- including expert preferences, holistic clinical tasks, safety judgments, and open-ended queries -- achieving 4 wins, 0 losses, and 18% lower variance. As a substantive finding, topic-level competency profiles expose striking domain-specific heterogeneity that aggregate accuracy masks. As a diagnostic tool, difficulty-tier analysis reveals two distinct response profiles (difficulty-sensitive responding and difficulty-insensitive responding) that require fundamentally different interventions. These results establish item-aware psychometric evaluation as a more valid and stable foundation for assessing LLMs in medicine, with potential implications for any high-stakes domain where benchmark integrity can be validated, and items vary meaningfully in difficulty and discrimination. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_24186 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Measuring Competency, Not Performance: Item-Aware Evaluation Across Medical Benchmarks Luo, Zhimeng Wu, Lixin Frisch, Adam He, Daqing Computation and Language Artificial Intelligence Accuracy-based evaluation of Large Language Models (LLMs) measures benchmark-specific performance rather than underlying medical competency: it treats all questions as equally informative, conflates model ability with item characteristics, and thereby produces rankings that vary with benchmark choice. To address this, we introduce MedIRT, a psychometric evaluation framework grounded in Item Response Theory (IRT) that (1) jointly models latent competency and item-level difficulty and discrimination, and (2) includes benchmark integrity validation to ensure items within each topic measure a single, coherent underlying ability. We prospectively evaluate 71 diverse LLMs on a USMLE-aligned benchmark across 11 medical topics. As internal validation, MedIRT correctly predicts held-out LLM responses on unseen questions with 83.3% accuracy. As external validation, IRT-based rankings outperform accuracy-based rankings across 6 independent external medical benchmarks -- including expert preferences, holistic clinical tasks, safety judgments, and open-ended queries -- achieving 4 wins, 0 losses, and 18% lower variance. As a substantive finding, topic-level competency profiles expose striking domain-specific heterogeneity that aggregate accuracy masks. As a diagnostic tool, difficulty-tier analysis reveals two distinct response profiles (difficulty-sensitive responding and difficulty-insensitive responding) that require fundamentally different interventions. These results establish item-aware psychometric evaluation as a more valid and stable foundation for assessing LLMs in medicine, with potential implications for any high-stakes domain where benchmark integrity can be validated, and items vary meaningfully in difficulty and discrimination. |
| title | Measuring Competency, Not Performance: Item-Aware Evaluation Across Medical Benchmarks |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2509.24186 |