Transparency in research: An analysis of ChatGPT usage acknowledgment by authors across disciplines and geographies

This study aims to systematically analyze the acknowledgment of generative AI tools, particularly ChatGPT, in academic publications. It delves into patterns across multiple dimensions, including geographical distribution, disciplinary affiliations, journals, and institutional representation.METHODOLOGY: Using a dataset from the Dimensions database consisting of 1,226 publications from November 2022 to July 2023, the study employs a variety of analytical techniques, including temporal analysis and distribution mapping across fields of research and geographical locations.FINDINGS: Acknowledgments are most frequent from authors affiliated with U.S. institutions, followed by significant contributions from China and India. Fields such as Biomedical and Clinical Sciences are highly represented, along with Information and Computing Sciences. Prominent journals like "The Lancet Digital Health" and preprint platforms including "bioRxiv" frequently feature acknowledgments, suggesting an accelerating role for AI tools in expediting research publication.LIMITATIONS: The scope of this study is restricted to the Dimensions database, possibly missing data from other platforms or non-indexed literature. Moreover, the study does not evaluate the quality or ethics of the acknowledgments.PRACTICAL IMPLICATIONS: The results are instructive for a range of stakeholders, including researchers, academic publishers, and institutions. They offer a foundational understanding that can inform future poli...
Source: Accountability in Research - Category: Medical Ethics Authors: Source Type: research