What type of study is this?

A multi-model longitudinal assessment of ChatGPT performance on medical residency examinations

View Full Paper

Ask AI

Bookmark

Authors

Multi-model evaluation shows ChatGPT models improve accuracy in medical residency exams, suggesting AI's role in education.

GPT-4o achieved 85.88% accuracy on medical residency exams, outperforming GPT-4 by a significant margin.
Both models displayed reduced accuracy on higher-order thinking questions, indicating room for improvement in critical thinking areas.
Research included assessments of 1,041 residency questions classified by Bloom's taxonomy, covering various medical areas.
Careful application of AI in education is essential due to ethical concerns and limitations in clinical practice.

Souto et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark