A new study has found that approximately one-quarter of medical papers published in 2020 were likely “made up or plagiarized,” underscoring a need to develop new tools to detect papers that corrupt scientific literature.
Using a so-called “fake-paper detector,” German neuropsychologist Bernhard Sabel found that out of 5,000 neuroscience papers published in 2020, 34% were likely false or plagiarized, according to a preprinted medRxiv study cited by Science on May 9. Similarly, the amount of “fake” medical papers published in 2020 amounted to 24%, the researchers found.
The culprit behind the allegedly large number of questionable materials finding their way into medical journals are paper mills, clandestine businesses that provide researchers with falsified papers or undeserved authorship for a price, according to the study. (RELATED: Princeton University Waits Six Months To Investigate Left-Wing Professor’s Alleged Plagiarism)
“Paper mills have made a fortune by basically attacking a system that has had no idea how to cope with this stuff,” Dorothy Bishop, a University of Oxford psychologist who studies fraudulent publishing practices, told Science.
Stunning new finding: Up to 34% of studies published in neuroscience journals, and about 24% in medical journals, are “made up or plagiarized.” https://t.co/rSYDkcxJmM
— David Boardman (@dlboardman) May 10, 2023
Sabel’s “fake-paper detector” detected questionable publications by examining the email addresses of the authors, specifically those who used private, non-institutional email addresses on their work. Through his efforts, Sabel was able to correctly flag nearly 90% of fraudulent papers in a test sample, according to the study results. Still, Sabel’s detection method was not without its issues, as it incorrectly targeted 44% of genuine papers.
Despite the imperfect solution, Sabel’s efforts and those of other researchers are making progress to eventually prevent the publication of falsified and plagiarized data, according to Science.
The International Association of Scientific, Technical and Medical Publishers (STM), an organization representing 120 publishers, is currently developing a new tool, known as the Integrity Hub, to help detect questionable submissions to scientific journals. The STM is keeping its detection method under wraps to avoid tipping off paper mills for work-arounds, but revealed they also hope to launch a separate tool that detects whether manuscripts have been simultaneously sent to more than one journal. Already considered unethical, the practice of multiple submissions is another indicator the manuscripts may have come from a paper mill, Science reported.
“It will never be a [fully] automated process,” Integrity Hub’s product director Joris van Rossum told the journal. The detection tools instead act more like “a spam filter … you still want to go through your spam filter every week,” he explained, in order to check for legitimate content that may have been flagged mistakenly.