The Effect of Using ChatGPT on Nursing Students’ Learning and Satisfaction in a Flipped Physiology Classroom

Masoumeh Gholami,1,2 Maryam Aalaa,1 Aeen Mohammadi11Department of Medical Education, School of Medicine, Tehran University of Medical Sciences, Tehran, Iran; 2Department of Physiology, School of Medicine, Arak University of Medical Sciences, Arak, IranCorrespondence: Maryam Aalaa, Department of Medical Education, School of Medicine, Tehran University of Medical Sciences, No. 2, Dolatshahi Alley, Naderi St, Keshavarz Blvd, Tehran, Iran, Email [email protected]: Nursing students in Iran are facing various challenges while learning complex physiological concepts. The flipped classroom approach can rarely be seen in the Iranian curriculum of nursing, and there is a lack of research on how to use AI-powered chatbots such as ChatGPT to offer the most appropriate nursing physiology education.Purpose: This study evaluated the effects of integrating ChatGPT into a flipped physiology classroom on nursing students’ learning outcomes and satisfaction.Methods: A pre-test/post-test intervention was conducted among nursing students. A total of 51 students were divided into an intervention (n=26) and a control group (n=25). Both groups were taught neurophysiology using a flipped classroom instruction, over four ADDIE model-based sessions. However, in the intervention group, the students were trained in prompt writing and asked to use ChatGPT to answer prompts in the presence of an instructor. The learning outcome and satisfaction of the students were evaluated using questionnaires. Data were analyzed in SPSS version 27, and P< 0.05 was considered statistically significant.Results: Following the intervention, the study group demonstrated significantly higher learning scores than the control group (7.15 ± 1.43 vs 5.52 ± 1.61; P < 0.001). Additionally, satisfaction scores were significantly higher in the intervention group (4.69 ± 0.35 vs 4.41 ± 0.49; P = 0.022). A non-parametric comparison of learning gain scores further confirmed a significantly greater improvement in the experimental group compared to the control group (U=132.50, Z=− 3.70, P< 0.001).Conclusion: A ChatGPT-supported flipped classroom improved learning outcomes and student satisfaction in nursing physiology education. These findings suggest that integrating artificial intelligence into medical education can enhance teaching effectiveness. Further studies with larger samples and other medical disciplines are recommended.Keywords: flipped classroom, large language models, artificial intelligence, physiology, nursing students, learning, satisfaction