Bytes versus brains: A comparative study of AI-generated feedback and human tutor feedback in medical education

Abstract

INTRODUCTION: Timely, high-quality feedback is vital in medical education but increasingly difficult due to rising student numbers and limited faculty. Artificial intelligence (AI) tools offer scalable solutions, yet limited research compares their effectiveness with traditional tutor feedback. This study examined the comparative effectiveness of AI-generated feedback versus human tutor feedback within the medical curriculum. METHODS: Second-year medical students (n = 108) received two sets of feedback on a written assignment, one from their tutor and one unedited response from ChatGPT. Students assessed each feedback using a structured online questionnaire focused on key feedback quality criteria. RESULTS: Eighty-five students (79%) completed the evaluation. Tutor feedback was rated significantly higher in clarity and understandability (p < 0.001), relevance (p < 0.001), actionability (p = 0.009), comprehensiveness (p = 0.001), accuracy and reliability (p = 0.003), and overall usefulness (p < 0.001). However, 62.3% of students indicated that both pieces of feedback complemented each other. Open-ended responses aligned with these quantitative findings. . CONCLUSION: Human tutors currently provide superior feedback in terms of clarity, relevance, and accuracy. Nonetheless, AI-generated feedback shows promise as a complementary tool. A hybrid feedback model integrating AI and human input could enhance the scalability and richness of feedback in medical education.

Ähnliche Arbeiten