Harnessing generative artificial intelligence for teaching statistics in medical research: Strategies for accurate hypothesis testing

Abstract

Generative artificial intelligence (AI) is reshaping the way statistics is taught and applied in medical research. This paper explores the use of AI tools, with a focus on Julius AI, to support hypothesis testing and statistical reasoning among medical students and researchers. Using a practical example and comparative analysis of prompting strategies, we demonstrate how AI can assist in performing t-tests and interpreting results. However, we also highlight the risks of misapplication and the importance of validating AI outputs. A structured framework is proposed to guide educators in integrating AI into the curriculum, emphasizing assumption testing, stepwise analysis, and critical evaluation. By embedding AI within a pedagogically sound approach, this work aims to enhance statistical literacy, promote responsible use of technology, and improve the quality of data analysis in medical education.