This analysis evaluates ChatGPT's accuracy on biology exam questions, highlighting concerns over academic integrity and response completeness.
In recent years, generative artificial intelligence (AI) tools, such as ChatGPT, have witnessed remarkable advancements, enabling their integration into classrooms by students. Nevertheless, incorporating such tools in education has raised concerns centered around issues such as academic integrity and the accuracy of statements made by ChatGPT. Investigating the accuracy of the content that ChatGPT produces is essential for informed decision-making and effective integration within educational contexts. In this study, we investigated how well ChatGPT performs on introductory biology exam questions. Our data source for this project was previous exam questions from an introductory biology course at a large, Southern, research-intensive university. Exam questions were coded into different levels of complexity, ranging from lower-order to higher-order thinking skills. ChatGPT explanations for each question were recorded and coded for both accuracy and completeness of the response. Our findings demonstrated variation in the accuracy and completeness of responses produced by ChatGPT by topic and Bloom's level.
No takes yet. Share an insight, caveat, or question.
Linnebur et al. (2025) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: