ChatGPT-Generated and Student-Written Historical Narratives: A Comparative Analysis

Abstract

This study investigates alternative approaches for demonstrating historical understanding in elementary school history education, motivated by challenges to educational institutions posed by increased ChatGPT-related plagiarism. Focused on secondary education, an area with scant research, this study, through sociocultural and linguistic methods of analysis, contrasted human-generated historical narratives with those produced by ChatGPT. It was found that ChatGPT’s narratives, while stylistically superior, lacked emotional depth, highlighting a key differentiation from human storytelling. However, despite this differentiation, ChatGPT otherwise effectively mimicked typical discourse patterns of historical storytelling, suggesting that narrative-based writing assignments do not significantly reduce the likelihood of ChatGPT-assisted plagiarism. The study concludes by suggesting that rather than focusing on mitigating plagiarism, educational approaches to ChatGPT should seek to channel its potential for historical narratives into assistance with task design, delivery of content, and coaching student writing.

Keywords

• history education
• historical understanding
• artificial intelligence
• writing
• secondary school
• assessment