Pooled LLM accuracy differs minimally by question source (71% for internal exams vs. 73% for public databases)¶
Claim · All claims
Evidence · 1 study · quant-synthesisr3·q2
Currency · newest study 2025 · not yet checked against its sources · not reviewed by a person · recheck by 2028: tool and model accuracy changes quickly
Subclaims¶
Reading the evidence codes
q is the design tier, i the impact, n the sample, and the kind with its r
the rigour judged within that kind: what each code means.
This claim: no standardized effect is recorded in observations/ for it yet, so no benchmark reading is possible.
q2 i? Subgroup analysis by question source showed only a minimal difference in accuracy (71% for internal exams vs. 73% for public databases, P=0.63), suggesting limited data-contamination effects. → Nouri Haniyeh 2025
Evidence¶
Nouri Haniyeh 2025¶
Nouri Haniyeh, Mahdavi Abdollah, Abedi Ali, Mohammadnia Alireza, Hamedan Mahnaz, Amanzadeh Masoud. (2025). Performance of large language models in medical licensing examinations: a systematic review and meta-analysis. Journal of Educational Evaluation for Health Professions. https://doi.org/10.3352/jeehp.2025.22.36
q2 · i? · quant-synthesis · r3
Subgroup analysis by question source reported in the Discussion as a check on potential training-data contamination from public question banks. The review reports "only a minimal difference in accuracy" between internal exams and public databases.
"our subgroup analysis based on question source showed only a minimal difference in accuracy (71% for internal exams vs. 73% for public databases), suggesting that this effect was limited."
Discussion¶
Related Claims¶
- Large language models achieve a pooled accuracy of 72% on medical licensing examination questions across 36 studies — related
- Subgroup analyses show significant between-model differences, with Copilot (75.25%) and ChatGPT-4 (72.97%) achieving the highest pooled accuracies — related
- ChatGPT-4.0 outperforms ChatGPT-3.5 in medical licensing exams, achieving a pooled accuracy of 81.8% versus 60.8% — related
- In in-training residency exams, ChatGPT-4.0 achieved 72.2% accuracy compared to 57.7% for ChatGPT-3.5 — related
- ChatGPT accuracy varies substantially across countries and specialties, from 73% on Italian to 22% on French examinations — related
Page metadata
| Field | Value |
|---|---|
status |
review |
generated |
by process:wiki-ingest on 2026-10-31 |
trust |
unverified — no one has confirmed this page's content |
Frontmatter is the page's machine-readable record — see the Schema & Guide.