Photographer:Fotograaf: Simone Golob
Three experts on the accusations of academic fraud against a professor
Last week, Observant reported about the Maastricht PhD candidate who accused a professor of Cardiology, her former supervisor, of academic fraud. Three experts consulted, all biologists who specialise in integrity matters, give their extensive views regarding this case.
What is the professor accused of?
The 29-year-old PhD candidate Eleni Liapi, born and raised in Greece, is doing research at UM-institute Carim into molecules and their role in heart failure. Back at the beginning of September, she approached the editors of Observant with accusations of “multiple image manipulations” by her former supervisor. We will not mention the name of this professor (yet); she only wishes to give a reaction when the Maastricht University Committee for Scientific Integrity, which is investigating the case, has passed judgement. According to Liapi, the case concerns a “pattern of misconduct” – but what exactly is she accusing the professor of doing?
The first allegation dates back to October 2017, when Liapi had been working there for more than a year. The professor had asked Liapi for her data, to be used for a subsidy application with the European Research Council. However, she noticed that the professor had changed the original labels on her data: from kidney cells to heart cells. A mistake, the professor reckoned, according to Liapi. And the professor partly reversed her ‘carelessness’.
The second discovery, in September 2018, was also put down to being “an unintended mistake” by the professor. Again, the case concerned a subsidy application, this time with ZonMW, a Dutch financer of health research. The professor supposedly changed the name of a regular biomedical technique into a more labour intensive and difficult method, while Liapi was certain that this could not be the case, because she had supervised the experiment herself. The professor promised to make the appropriate corrections should the application be approved, but it was not approved.
The third, and according to Liapi the most serious manipulation, dates back to June 2019. In a PhD thesis supervised by the professor, which had shortly before been defended, an original picture of a blot (a biomedical technique that can be used to find a specific protein) from 2015 had been reused. But this time with ‘adapted’ labels. Liapi is certain that it has been tampered with.
What do the experts say?
Observant put the accusations before eight foreign experts. A response came from three of them. The first is molecular biologist Rune Linding, researcher at Humboldt Universität in Berlin and leader of the Linding Lab. The second is Leonid Schneider, a German cell biologist, now a research journalist who deals with integrity queries. The third is Elisabeth Bik, microbiologist, who obtained a PhD in Utrecht and who works as a science consultant in the United States and is an expert in the field of the fraudulent use of images.
Observant’s question focuses in particular on the last accusation, the one of reusing an image from 2015. Is this actually a case of duplication? Or do the two images simply look alike and there is nothing wrong?
“This is data manipulation, definitely intentional,” Leonid Schneider reckons. He is very detailed in his e-mail. “Obviously, someone saw the original labelling and decided to change (or fake) it to fit some preconceived narrative. The results from this experiment are not reliable anymore, and the general scientific ethics of the scientists responsible are in question now.”
According to Rune Linding, this case could theoretically be “extreme carelessness” but it appears to be more like “misconduct” because it seems systematic to him. “If there is documentation, it should be possible to unravel what happened. Changing labels strongly suggests intent.” He claims with “almost 100 per cent certainty” that the image from the thesis is a duplicate. “But one would need to run image forensics to properly conclude this”, he says. “Science is full of mistakes and bad science, but it’s also supposed to self-correct this. What science is very sensitive to, is bad intent: fraud, manipulation, misconduct. If such ‘science’ is published or used to obtain research funding, not only can it mislead many scientists and distort the scientific record, it can even hurt patients.”
Finally Elisabeth Bik. According to her, the original image from 2015 and the image in the thesis look very similar, “but they differ in the number of hairs and stains.” In answer to the question how she would qualify this behaviour, she said: “Misconduct can only be determined if the institute investigates the history of this blot, and what has been done to it. I cannot determine that by just looking at the images.” But should an investigation prove the PhD candidate right, then as far as Bik is concerned, this is not a matter of sloppy labelling; “it involves actively changing the labels of the figures, to make them represent something other than what they originally were.”
Rune Linding calls for an independent investigation, and no, not by a committee at Maastricht University, but by an external party. “Many (most) universities try to cover up research misconduct, only when external parties or the public gets involved, will they try to deal with it.”
With assistance from Maurice Timmermans
PhD candidate Eleni Liapi already lodged her formal complaint with the Committee for Scientific Integrity (CSI) in July 2019. The latter only issued an admissibility statement after 42 days, longer than the official period mentioned in the rules and regulations.
The advice also takes longer than the required twelve weeks. According to UM spokesperson Fons Elbersen, the reason, among others, is the complexity of the case. The CSI decided to consult an external expert who wishes to stay anonymous, Elbersen says. "UM respects his/her wish."
The advice of the CSI is still not on the Executive Board’s desk, he adds. But it's up to them to take a decision afterwards, within four weeks, and inform the ‘parties’ in writing.
In answer to the question whether the Board will share its decision with the rest of the world, Elbersen says: “That is not common practice.” But in this case, which has already been picked up by the media, he says the Board has yet to decide.
Should Liapi not agree with the verdict, she has the right to appeal to the national body of Scientific Integrity, (LOWI, het landelijke orgaan Wetenschappelijke Integriteit). The latter organisation will give (a non-binding) advice to the Executive Board about the decision.
The outcome will be made public (in anonymised fashion) on the VSNU website after the procedure has been completed.
Parallel to the integrity procedure, there is also an ongoing labour dispute, now lodged with the court of justice. Liapi had to hand over the keys to the lab where she worked to her former supervisor in February 2019, which meant that she could not continue her experiments. The judge will cast a verdict on the matter next week, unless a solution is reached sooner.