Evaluation of the applicability of ChatGPT in biological nursing science education

Abstract

Purpose: The purpose of this study was to evaluate the applicability of ChatGPT in biological nursing science education. Methods: This study was conducted by entering questions about the field of biological nursing science into ChatGPT versions GPT-3.5 and GPT-4 and evaluating the answers. Three questions each related to microbiology and pharmacology were entered, and the generated content was analyzed to determine its applicability to the field of biological nursing science. The questions were of a level that could be presented to nursing students as written test questions. Results: The answers generated in English had 100.0% accuracy in both GPT-3.5 and GPT-4. For the sentences generated in Korean, the accuracy rate of GPT-3.5 was 62.7%, and that of GPT-4 was 100.0%. The total number of Korean sentences in GPT-3.5 was 51, while the total number of Korean sentences in GPT-4 was 68. Likewise, the total number of English sentences in GPT-3.5 was 70, while the total number of English sentences in GPT-4 was 75. This showed that even for the same Korean or English question, GPT-4 tended to be more detailed than GPT-3.5. Conclusion: This study confirmed the advantages of ChatGPT as a tool to improve understanding of various complex concepts in the field of biological nursing science. However, as the answers were based on data collected up to 2021, a guideline reflecting the most up-to-date information is needed. Further research is needed to develop a reliable and valid scale to evaluate ChatGPT’s responses.

Ähnliche Arbeiten