AYLA İREM AYDIN, DOĞUKAN REİS
Bandırma Onyedi Eylül Üniversitesi Sağlık Bilimleri ve Araştırmaları Dergisi - 2025;7(1):73-79
Aim: The aim of this study was to compare the success performances of ChatGPT models and student nurses in the course of nursing approach in pediatric emergencies. Materials and Method: This study was designed as a retrospective and comparative analysis. The study was conducted by using the end-of-semester exam questions of the nursing approach in pediatric emergencies course in the nursing department of a state university. In the exam, 40 multiple-choice questions with five options were directed to ChatGPT 3.5 and ChatGPT 4.0 models. The answers given to the questions were compared with the answers given by the artificial intelligence robots and the averages of the answers given by the student nurses taking the course. Results: ChatGPT 4.0 gave 90% correct answers to the questions, while the students gave 82.32% correct answers. The lowest correct response rate was 55% for ChatGPT 3.5. In this exam, ChatGPT 4.0, students and ChatGPT 3.5 achieved the highest success respectively (p < 0.05). Conclusion: ChatGPT 4.0 performed well with the correct suggestions it provided on nursing approach in pediatric emergencies. Artificial intelligence-based systems can be used in nursing education as a reference tool to assist the instructor and students of the course. It is important to develop strategies and conduct studies on how these technologies can be used more effectively in nursing education.
