ChatGPT performance on scenario-based multiple-choice questions (MCQs) in medical physiology

Abstract

Abstract Background and objective: With the rapid advancement of artificial intelligence, natural language processing models like ChatGPT have gained prominence in various educational domains. This study investigated the performance of ChatGPT in answering scenario-based multiple-choice questions (MCQs) within the intricate field of medical physiology. Materials and methods A comprehensive dataset of 50 scenario based MCQs covering cardiovascular system, respiratory system, Endocrinology, Renal system, central nervous system topics in medical physiology was compiled. ChatGPT was then tested with these questions, and its responses were evaluated based on accuracy and contextual appropriateness. The first response obtained was taken as the final answer for scoring and analysis. If the correct option was chosen by ChatGPT, 1 mark was awarded, and the wrong option was awarded zero mark. Results Out of 50 MCQs in various topics in medical physiology, ChatGPT attempted all the MCQs, and total score obtained is 30/50 (60%) marks. ChatGPT performed most accurately in respiratory system MCQs with 70% (7/10) and least accurately in renal system MCQs 50% (5/10). Conclusion ChatGPT attempted the scenario based MCQs and attained a satisfactory score.

Ähnliche Arbeiten