Study reveals extent of ‘science scam’ in academic publications
Every fifth article in the world, and 2.4% of articles in Russia are fake.
German researchers have found that about 20% of articles published in scientific journals may contain fake data created by illegal “fake article factories” that are paid for falsifying scientific papers. The study confirms growing evidence that academic publishers are facing an explosion of fake research being sold to academics in need of publications to advance their careers. Scientists believe that most of the fake research comes from China.
A team led by Professor Bernhard Sabel, head of the Institute for Medical Psychology at the Otto von Guericke University in Magdeburg, found that the number of fake articles has increased significantly in recent years. Fake news is especially popular in China, where some hospitals and health authorities require doctors to be the first authors of a certain number of articles. Reviews of clinical data lose credibility when fake studies are included, undermining public confidence in science and medicine. The science sector in China is also suffering from Western perceptions that Chinese researchers have a carefree attitude towards the integrity of published work. “The forgery of scientific publications is perhaps the biggest scientific hoax of all time, wasting financial resources, slowing down medical progress and possibly threatening lives,” Zabel said.
Fake biomedical research is becoming more common, especially in the neurosciencese.
Most independent researchers who track scientific deceit analyze the content of articles and look for, for example, fake or stolen images or certain text sequences. Academic publishers are also beginning to adopt more sophisticated deception detection tools. The German researchers went the other way, identifying simple red flags that do not require detailed study of the article itself, such as the use of private rather than institutional email addresses, hospital affiliation rather than university, and the absence of international co-authors.
Global contribution to fake publications (%).
These factors were confirmed by comparing a sample of known forgeries with articles believed to be genuine. An article that was published in as a preprint on MedRxiv , emphasizes that a red flag is not a definitive indication of a hoax, as it can falsely identify a significant number of genuine articles. The number of red flag publications across biomedicine has grown from 16% in 2010 to 28% in 2020, with a much steeper increase in neuroscience than in clinical medicine. Considering articles flagged as fake that are actually genuine, Zabel estimated that the actual percentage is now around 20%, equivalent to about 300,000 articles a year.
Referring to the “mass production” of fake studies, the scientists also studied the methods used by fake factories, whose annual revenues are estimated at $3-4 billion. “They typically use sophisticated AI-based text generation technologies, data and statistics manipulation and falsification, image and text piracy,” the study says. Professor Gerd Gigerenzer of the University of Potsdam, a psychologist and co-author of the article, said: “It will be a race between fake article factories and those of us who try to detect them, with both sides using artificial intelligence.”
But the ultimate solution, he said, is to ease the pressure to publish, especially in China. Others, he thinks, could follow the example of the German Science Foundation, which tells applicants for funding that they must limit their own papers to five.
Forgery of scientific articles is not only an ethical problem, but also a threat to science and society. Fake articles can mislead other scientists, influence policy and practice, and undermine the credibility of the scientific community. Unfortunately, this problem is becoming more common and complex as counterfeit factories use modern technology to create and distribute counterfeit works.
Combating this phenomenon requires concerted action by all stakeholders: scientists, publishers, editors, reviewers, science funders, and governments. It is necessary to develop and apply effective counterfeit detection tools, such as image, text and data analysis. It is also necessary to raise awareness and responsibility of participants in the scientific process about the risks and consequences of forging scientific articles. In addition, it is necessary to reduce the pressure on publication and create conditions for honest and high-quality scientific activity.
Forgery of scientific articles is a serious problem that requires immediate attention and solution. This is not only a matter of professional ethics, but also of the public good.