Key points are not available for this paper at this time.
High Resolution Image Download MS PowerPoint Slide With recent technological advancements, there are concerns related to the use of generative artificial intelligence in academia by students and faculty. We aim to spark conversations surrounding the best practices for using artificial intelligence in chemistry education research. To this end, we asked ChatGPT to analyze a dataset we previously analyzed and published in this Journal . In the previous project, we analyzed the academic integrity statements in chemistry course syllabi and found that syllabi were largely consequence focused, emphasizing the need for faculty to clearly outline expectations regarding what counts as cheating. In the present article, we explore ChatGPT’s ability to engage in qualitative research, focusing on considerations related to building a case for trustworthiness, including how it independently coded qualitative data and how it engaged with inter-rater reliability. As part of this, we also asked ChatGPT to compare our analysis with its analysis. Regardless of how well ChatGPT can perform research tasks, we maintain the importance of transparency and thorough detail when disseminating a project that involves AI-assisted research. To this end, we highlight the supporting role of AI in research─to quote ChatGPT, “AI as a research partner, not a replacement”.
McAfee et al. (Mon,) studied this question.