Today, the University of Surrey has revealed shocking trends on the production of pattern-based research papers. This problem is particularly acute among studies that use NHANES data. A systematic literature search conducted by a team led by post-graduate researcher Tulsi Suchak analyzed studies published over the past decade, uncovering a significant increase in questionable statistical methodologies and study designs.
The study’s research team reviewed literature over the past decade, pinpointing 341 reports released across 192 journals. The amount of research that has come out since then has skyrocketed. From 2014 to 2021, the papers were written by researchers at an average of four papers a year. This number jumped to 33 in 2022 and is on track to surpass 82 in 2023. In addition, the first ten months of 2024 already produced a shocking 190 publications.
Tulsi Suchak called for greater transparency about what data is being used for what research. “We’re not trying to block access to data or stop people using AI in their research – we’re asking for some common sense checks,” she stated. After hearing Suchak speak, it’s easy to understand why there’s a rising tide of concern among the academic community that we’re publishing an integrity of research.
The evidence from the study suggests most manuscripts may look scientific, but have little to no foundation under a proper microscope. Matt Spick is a senior lecturer in health and biomedical data analytics at the University of Surrey. As a co-author of the report and someone deeply concerned about these trends, Mr. He remarked, “We’ve seen a surge in papers that look scientific but don’t hold up under scrutiny – this is ‘science fiction’ using national health datasets to masquerade as science fact.”
The report also called attention to the increasing corporate authorship affiliations observed in recent years. Between 2014 and 2020, the primary authors on just two of these 25 manuscripts had affiliations with institutions based in China. The percentage from 2021-2024, however, increased to a staggering 292 of 316 manuscripts.
The study argued for including reviewers who possess appropriate expertise to address these issues. In particular, it recommended creating an “Orange Light” to flag studies that just scratch the surface of larger health issues. The authors hope that these measures will prevent subpar science from slipping through the cracks of the peer-review process. Third, they’ll boost the reproducibility and quality of scientific research.
The boom in predatory publications has brought more light to the larger issues facing academic journals. This phenomenon echoes previous concerns raised when Wiley discontinued 19 scientific journals managed by its Hindawi subsidiary last year due to similar issues.
“The use of these easily accessible datasets via APIs, combined with large language models, is overwhelming some journals and peer reviewers, reducing their ability to assess more meaningful research and ultimately weakening the quality of science overall,” Spick commented, summarizing the implications of these findings.
The University of Surrey’s research highlights an important turning point for the future of scientific publishing. It calls for action to ensure that research maintains its integrity and relevance in an era where data accessibility and artificial intelligence are reshaping traditional methodologies.
