South KoreaEducation2025 – ongoingGachon University
AI simulated patients for clinical-examination preparation
A Gachon-led pilot involving 19 students found no significant examination-score difference between AI practice and peer role-play
Problem
Clinical interview practice depended on peers or simulated patients. The researchers tested whether an AI patient could provide practice and feedback for Korean medicine students.
Approach
The team built a chatbot combining GPT-4o and Claude 3.5 Sonnet. In a randomized pilot, students used either the chatbot or peer role-play before clinical examinations. The chatbot generated patient responses, scored submitted interviews and provided written feedback.
Results
Takeaway
The small pilot did not establish that AI practice improved examination performance over peer role-play. It supported further evaluation rather than a claim of educational superiority.
Categories
- Industry
- Education
- Business function
- Production and qualityIT and R&D
- Innovation type
- Process innovation
- AI technology
- Generative AIConversational and voice AI
- Adoption
- Built in-house
- Evidence
- Independent study
Sources
Source summaries
The peer-reviewed study compared AI interview practice with peer role-play in 19 students. Examination-score differences were not statistically significant.
Full details are in the original sources.