Using Artificial Intelligence to Evaluate the Nature of Science Myths: A Comparison of Pre-Service Science Teachers and ChatGPT

Abstract

This chapter aims to compare the performances of pre-service science teachers (PSTs) and generative artificial intelligence (AI) application (ChatGPT) in answering nature of science (NoS) myths. In this study, which was designed according to a qualitative descriptive research design, the 15 most frequently encountered NoS myths were addressed to 104 PSTs and the ChatGPT. The data were analyzed using content analysis. As a result of the research, when the items related to all myths were analyzed, approximately 48% of the PSTs had myths about the NoS, while 66% of the ChatGPT answers included correct information. Therefore, ChatGPT can be effective in seeking answers to NoS myths. However, since the resources used by ChatGPT are limited to certain books, contents, or resources of certain countries, it may not be efficient enough in terms of providing information diversity. According to these results, comparison of PSTs and ChatGPT's answers are discussed.