Study reveals scale of ‘science scam’ in academic publishing

One in five articles published in journals may contain faked data produced by unauthorised “paper mills” that are paid to fabricate scientific submissions, according to a study by German researchers who used new techniques to “red flag” problematic papers.

The study adds to the growing evidence that academic publishing faces a damaging surge in fabricated research sold by paper mills to researchers desperate for published work to boost their careers. It also backs up recent evidence that the majority of fake research comes from China.

The team, led by Professor Bernhard Sabel, who heads the Institute of Medical Psychology at Otto von Guericke University Magdeburg, found that the number of fake papers had risen substantially in recent years. Pressure to publish had been particularly intense in China, they said; for example, some Chinese hospitals and health authorities require physicians to be first author on a set number of papers.

Reviews of clinical evidence lose credibility when fraudulent studies are included, undermining public trust in science and medicine. China’s science sector also suffers from the western perception that the country’s researchers have a cavalier attitude to the integrity of published work.

“Fake science publishing is possibly the biggest science scam of all time, wasting financial resources, slowing down medical progress and possibly endangering lives,” said Sabel.

You are seeing a snapshot of an interactive graphic. This is most likely due to being offline or JavaScript being disabled in your browser.

Most of the growing band of independent investigators who track scientific fraud analyse the content of papers and look, for example, for manipulated images and implausible genetic sequences. Academic publishers are also beginning to adopt more sophisticated fraud detection tools.

The German researchers took a different tack, identifying simple “red flag” indicators that do not require detailed examination of the paper itself, such as the use of private rather than institutional email addresses, affiliation with a hospital rather than university and lack of international co-authors. These were validated by comparing a sample of known fakes with papers regarded as genuine.

The paper, which has been posted as a preprint on MedRxiv but has not been peer reviewed, emphasises that the red flag is not a definitive indication of fraud, because it can falsely identify a substantial number of genuine papers.

The number of red flag publications across biomedicine rose from 16 per cent in 2010 to 28 per cent in 2020, with a much sharper increase in neuroscience than in clinical medicine. Taking account of papers flagged as fake that are actually genuine, Sabel estimated that the actual proportion now was about 20 per cent, equivalent to around 300,000 papers a year.

You are seeing a snapshot of an interactive graphic. This is most likely due to being offline or JavaScript being disabled in your browser.

Citing the “mass production” of faked research by paper mills, the researchers also investigated the techniques used by a sector whose annual revenues were estimated at $3bn-$4bn. “They typically appear to use sophisticated AI-supported text generation, data and statistical manipulation and fabrication technologies, image and text pirating,” they said.

Professor Gerd Gigerenzer of the University of Potsdam, a psychologist and co-author of the paper, said: “It will be a race between the paper mills and those of us who try to detect them, with both sides using AI.”

But the ultimate solution, Gigerenzer added, was to reduce the pressure to publish, particularly in China. Others, he suggested, could follow the example of the German Research Foundation, which tells applicants for funding that they should limit the number of their own papers cited to five.

Jennifer Byrne, an oncology professor at the University of New South Wales and leading sleuth, who was not involved in the project, said: “It’s an important study because very few studies have been published on this large scale. It is pointing to a massive problem.”