This study aims to examine students’ responses to information generated by ChatGPT, based on the premise that examining fundamental nature and characteristics of the technology is a prerequisite to its effective utilization.
To this end, this study first identified the characteristics of historical information generated by ChatGPT. The historical information produced by ChatGPT is not presented as fragmented data, but as a well-structured narratives or writings resulting from searching and processing materials. It also bears similarities to the historical information encountered by students in that it includes multiple layers of false information created by AI hallucinations. However, ChatGPT-generated information differs from conventional historical information in that it does not provide sources and lacks the intervention of an author’s intent.
Subsequently, documents containing historical information were generated using ChatGPT, and the types of errors in these documents were examined. Three groups were then surveyed to understand how they perceived and responded to the false information in the ChatGPT-generated document.
As a result, it was confirmed that knowing historical facts, including the process of knowledge construction, is important and that lessons need to shift toward fostering critical thinking and a questioning mindset.
Based on these findings, it was concluded that in order to use ChatGPT in classroom instructions effectively, it is necessary to provide guidance on the technology’s limitations and potential errors, combine historical facts with a critical perspective, verify the process of constructing historical facts and knowledge within generated information, and examine the context and logical structure of the writing. Additionally, this study suggested a broad outline of an approach for utilizing ChatGPT in history classes.