Scientific publication fraud is becoming more widespread, warns a US study

This is a study in the form of a cry of alarm that a multidisciplinary team of researchers published on Monday, August 4, in PNAS, the journal of the National Academy of Sciences of the United States. "Scientific fraud is growing much faster than scientific production as a whole," we read in this article, whose first author, Reese Richardson, is a researcher in computational biology at Northwestern University in Illinois.

For this massive undertaking, these mathematicians and other biologists, for example, sifted through 276,956 articles published between 2006 and the end of 2023 in PLOS One , a peer-reviewed online journal, and tracked 134,983 authors and 18,329 reviewing editors who validated these articles. This American journal was chosen not because of any specific suspicions, but because the metadata associated with the articles it publishes is transparent and usable.

What did they find? That some of the filters that were supposed to guarantee the peer review process, a principle at the heart of trust in scientific publishing, were flawed. Or worse, were being corrupted by fraudulent practices.

The study cites 45 PLOS One editors who had an unusually high rate of accepted publications that were later retracted or criticized on PubPeer , the leading site for post-publication peer review. These editors (0.25% of all editors in the journal) "edited 1.3% of all articles published in PLOS One , but 30.2% of the retracted articles." Compounding the problem, more than half of them also authored articles published by PLOS One that were later retracted. The most common reason for article retraction is the authors' professional misconduct (such as falsified data or plagiarism), or the editors' misconduct, such as the lack of minimal quality control of submitted work.

You have 71.16% of this article left to read. The rest is reserved for subscribers.