“The integrity of science is important. It must be credible. Every new study is based on existing studies – if these are wrong, the research continues in the wrong direction and eventually the whole thing becomes useless,” says Lonni Besançon.
He is an assistant professor at the Department of Science and Technology, where he explores how data can be visualised and used in areas such as healthcare and the judiciary. But in addition to his own research, he also reviews other researchers’ works after they are published.
This is called academic sleuthing. A job that is both thankless and unpaid.
“No one thanks you for finding something bad. Besides, it’s not part of my contract. No one is employed to check scientific integrity compliance after publication, but this is something I, and others like me, do outside of working hours,” says Lonni Besançon.
Fabricated data
But what do academic sleuths check? To understand this, we need some background:
The process of getting a scientific article published in a journal can be broken down into a number of steps. Once the study is finished and the data is collected, the researchers write a draft, or manuscript, of an article. That manuscript is then sent to one or more publishers in the hope that it will be accepted by a journal. For the article to be accepted, it must go through something called a peer review. This is a kind of review that involves other researchers in the field reviewing the article to see that it is of good scientific quality. Hopefully, the article will then be published. It is a process that often takes a long time, in some cases several years.
“Peer review is based on mutual trust. When I read someone else’s article, I take it for granted that no-one is cheating. In my field, there’s less risk as it’s a relatively small field and we all know each other. But in microbiology, for example, where there are hundreds of thousands of scientists who will never meet each other, the risk of cheating increases.”
In academia, cheating can mean anything from removing data points to get the results you want, to lack of ethical reviews and completely fabricated data and results featuring AI-written articles. And the fact that some such articles still slip through the peer review process is a major concern.
Incentives for cheating
Universities, other higher education institutions and individual researchers pay large sums of money to publishers such as Springer, Elsevier and Nature in order to publish and for the review to be correct. But often the underlying code or data is not checked, only the article itself. And for smaller publishers, it is not even certain that there will be any regular peer review.
“It’s completely mad! We’ve found articles that were submitted, accepted and published on the same date. This sets off warning bells, something isn’t right.”
According to Lonni Besançon, the widespread cheating is due to a fundamental error in how the academic process is structured. The number of publications often plays a greater role than which journal the researcher has published in when it comes to obtaining a high H-index. This is an index intended to show how productive and skilled a researcher is. It may form the basis for future employment and, above all, research grants.
“Publishing a zero result is almost impossible, although it is equally important for the research community to know that something doesn’t work. But if you delete some data points, it is no longer a zero result, and all of a sudden publishable. So the incentives for cheating are built into the academic process.”
Low retraction rate
When Lonni Besançon and his colleagues find something wrong in a scientific article, they write to the publisher and point it out. Most often, they receive no reply at all. Sometimes the publisher replies that they have received the complaint but then does not take it further. And in rare cases, the scientific article is actually withdrawn. But the academic sleuths are not credited for this. Often it says just that the magazine has withdrawn the article and nothing more.
One of the most recognised and successful academic sleuths is Elisabeth Bik, a microbiologist from the Netherlands. She and Lonni Besançon have worked together several times,and she is a great role model for him. Elisabeth Bik has reported approximately 8,000 incorrect scientific articles but less than 20 percent have been removed by the journals. A disappointing figure.
The work that Lonni Besançon has received the most attention for concerns research fraud at a French research institute. They published false results at the beginning of the COVID-19 pandemic that claimed they had found a cure. But something was not right.
“I’m not an expert in biology or disease. But statistics is something I really know and there was something very wrong in their articles on several points.”
Harassment and defamation
Among other things, he could see that the same ethical review number was used in 248 studies when there should actually be a unique number for each study. The studies were done in many different ways – samples of blood, saliva, faeces and skin were collected from both young and old study participants, including children. This should require several different ethical reviews.
When Lonni Besençon and his colleagues presented their findings in the journal Research Integrity and Peer Review, this caused a ferocious debate, especially in France, about research ethics and he appeared on national television, newspapers and radio. The news also spread across the English Channel and The Guardian published a long article.
But that was also when the storm began for Lonni Besançon.
“They write about me on Twitter all the time and have also emailed all my colleagues here and tried to get me fired. They smear my name and harass me in different ways. I’ve also had a couple of calls with death threats.”
How do you cope with that?
“If they see me as a target, it means I’ve done something right.”
The scientific journal Science has published an extensive article about Lonni Besançon and his sleuthing colleagues. The article also provides further insight into the most notable case.
It can be read here: ‘Failure at every level’: How science sleuths exposed massive ethics violations at a famed French institute
The article: Raising concerns on questionable ethics approvals – a case study of 456 trials from the Institut Hospitalo-Universitaire Méditerranée Infection; Fabrice Frank, Nans Florens, Gideon Meyerowitz-katz, Jérôme Barriere, Éric Billy, Véronique Saada, Alexander Samuel, Jacques Robert, Lonni Besançon; Research Integrity and Peer Review volume 8, Article number: 9 2023; published online 3 August 2023. DOI: 10.1186/s41073-023-00134-4
