Biologist Stan van Pelt: ‘Science should not be a competition to see who has the highest publication score or who discovers something first.’ PHOTO: PAUL RAPP
Errors, sloppiness, fraud, conflicts of interest, corruption, deception: science has it all. Biologist and science journalist Stan van Pelt sounds the alarm: ‘Fraud is a growing but underestimated problem.’
‘There is work to be done. Public trust in science is high, and rightly so: most researchers do their work with integrity and conscientiousness. But there is a great deal wrong, and researchers are only partly aware of it’, says science journalist Stan van Pelt. He has described all manner of fraud – from careless mistakes to criminal practices – in the book Sloppy Science, which was published on 24 September. The conclusion: quality control is falling short. Fixing that will require an overhaul of the academic system.
Fake paper
Van Pelt studied biology and worked as a neuroscientist until 2017, when he moved into science journalism and wrote extensively about research integrity. ‘There are several aspects to that. This book gave me the space to address all of them.’ He immersed himself in the subject – literally. He not only spoke with those involved in individual cases of fraud, such as whistleblowers, researchers who had ‘fallen from grace’, and departmental boards, but also took part in organised fraudulent practices by, entirely openly, publishing a fake paper and attending a fake conference. ‘That way I could better explain what these practices, which threaten science from the outside, actually involve. This was the most enjoyable part to do.’
ChatGPT
He submitted a paper to the online journal Cases on the brain activity of people exchanging information telepathically. To make the nonsense complete, he added a sub-experiment purporting to demonstrate interference by extraterrestrial beings. ‘I made up the experiments and results using AI; ChatGPT wrote the text and DALL-E produced the figures. It was done in an hour.’ Since 4 February it has been available online, looking for all the world like a perfectly ordinary article. But the discussion section and the methods statement make clear that it is fake, and the axes on the figures are illegible. ‘It shows that no editor or reviewer so much as glanced at it. Predatory journals such as Cases parasitise the open-access revenue model of publishers. They invite researchers with flattering words to publish something, charge a substantial fee, but skip the quality control they have been paid to provide.’ Van Pelt himself did not have to pay on this occasion, on the grounds that his publication budget had supposedly run out.
Paper mills
What he can do, others can too. ‘There are paper mills – article factories – where you can order a paper. It is either entirely fabricated or plagiarised from other articles. Predatory journals publish it, but sometimes legitimate journals do too. It happens far more than I realised; it is a colossal problem. It is also possible to buy co-authorship of an article shortly before it is published. This is openly advertised on platforms such as Telegram.’
Since 4 February it has been available online, looking for all the world like a perfectly ordinary article
Equally brazen are invitations to fake conferences. Van Pelt attended an edition of the NeuroTalk conference, to which he had been invited by e-mail as a former neuroscientist. Rather than paying the steep admission fee, he booked a room in the hotel where it was being held and simply walked into the sessions. ‘The conference does actually take place, and real scientists from real universities do give talks. But once you are there, you can see that something is not right. Of the grand announcement with its many parallel sessions and occasionally impressive names, little remains. There are very few attendees – only the people presenting. They find themselves in a session with a random assortment of topics, delivering their talk to an audience that has no particular interest in it, with no productive discussions and no valuable contacts made. Afterwards, participants feel cheated. That was a one-off – never again, they told me, almost without exception.’
When he confronts the organisers with his findings, they respond angrily and even threaten legal action. ‘And yet, on a different e-mail address, I received an invitation to the next edition of the conference.’
Temptation
Fraud, fake papers, predatory conferences: the temptation to get involved can be considerable for those under performance pressure or seeking to boost their status as a scientist. That is why Van Pelt wants to move away from the metaphor of science as elite sport. ‘Because that is not what it is. It should not be a competition over who has the highest publication score or who discovers something first. What matters is that researchers build knowledge together.’ The book offers recommendations for making the atmosphere in academia less competitive. Publishers, too, must take their responsibility. ‘I have great respect for so-called fraud detectives such as Elisabeth Bik and Ben Willem Mol, who scour publications for manipulated images and statistical flaws. Quite a number of dubious papers have been retracted or flagged with a warning as a result of their efforts. They are remarkable people, who understand that every instance of fraud reflects on science as a whole and feel a sense of responsibility. Working as sole operators, they put in an enormous amount of work and receive nothing but grief in return – up to and including threats of legal action. What they are doing is really the publishers’ job. Publishers earn enough money from science to be able to deploy more people for quality control.’
SLOPPY SCIENCE. ALLES OVER WETENSCHAPSFRAUDE VAN FOUTJES TOT NEPCONGRESSEN – Stan van Pelt
Paperback, 304 pages, €25 (e-book: €13)
PHOTO: PAUL RAPP