And Lead Us Not Into Temptation
Fraud scandals shake the credibility of science. The perpetrators are driven by personal motives such as the craving for recognition or greed for success. But the real reasons for fraud and manipulation are embedded in the system.
Curiosity and greed are dangerous bedfellows. A scientist submits a funding application although he already knows the results. Another publishes a manuscript that he was supposed to be reviewing as his own work. A research group leader reports that publications written by members of her group are being accused of being manipulated. And, indeed, the young scientists actually had falsified the data. But worse than that, their boss had taken no interest in the authenticity of the findings and had even been cited as co-author.
These are all cases that have happened in Germany. The German Research Foundation (Deutsche Forschungsgemeinschaft, DFG) has compiled the information and knows the names of everyone involved. Manipulation is an issue that now reaches up into the highest institutions. It even affected Germany’s most important research award, the Alexander von Humboldt Professorship. In 2015, the social psychologist Jens Förster of Ruhr-Universität Bochum relinquished the Humboldt Foundation’s five million euros award after his former university had raised allegations of data manipulation.
Greed rots the mind
The Council is critical of the fact that there is a growing trend “to present one’s results as groundbreaking in order to be able to publish them”. Biomedical research, in particular, promises much. “Greed rots the mind,” says Wolfgang Löwer, the former German Research Ombudsman, who has been dealing with cases of academic misconduct for more than ten years. Löwer shudders when he thinks of a paper in the journal Nature which even made it onto the evening news in 2014. The experiment: you take mouse cells, pour citric acid over them, mix it all up and – done. This is the way to turn around the ageing process, to create stem cells. Therapies for millions of people. The salvation of humanity. But the Japanese-US research group had cheated – the article was retracted.
That there were no German researchers involved in this case is pure chance, because research is no longer a national affair. It is internationally connected. The mechanisms that lead to fraud, deception or to looking the other way are universal. “The competition is huge. And, on top of this, today everyone works in teams. That makes it easier to cover up who is responsible for what, which can favour deception,” says Löwer. In one edition of Nature there was an article on the decay of elementary particles citing no fewer than 3,000 contributors. Such extreme examples illustrate how difficult it is to prove responsibility when misconduct is discovered.
“You take mouse cells, pour citric acid over them, mix it all up and – done.”
Of course, most researchers do work responsibly. It is not as though Löwer deals with thousands or even just hundreds of such cases every year. But that is not saying much, because he is only concerned with cases that are not dealt with by the universities themselves. There are now ombudspeople in many places in Germany. But there is no data base collating the cases. And this grey area is difficult to assess, according to Löwer. “We can only follow up on things that are reported to us.” Usually, the people who are able to report on misconduct are part of the system themselves. Putting one’s own reputation on the line, harming one’s own career, endangering the future of the laboratory and the colleagues working in it is a threshold that is difficult to cross.
Who is going to admit to a craving for recognition?
All disciplines are susceptible. The form of misconduct made famous by prominent examples – plagiarism – is, however, on the decline, according to Dorothee Dzwonnek, DFG Secretary General. Taking a look at the cases she has investigated, it is fraud in the lab and amongst researchers who work on their own data that is on the up. “The problem is accelerating in the life sciences, in particular. Pressure increases and with it the danger, and sometimes the tendency, that people consciously or unconsciously make mistakes.” Anxiety about the future is the main reason given in the explanations of people’s conduct. Who is going to admit to a craving for recognition?
In a series of articles in the journal The Lancet in 2014, the authors claimed that 80 per cent of biomedical research was “waste”. A result that cannot be replicated, even if the conditions of the original experiment are reproduced precisely, is also “waste”. “It’s only true if you can cook it for yourself,” says Löwer pithily. This is the core principle of science and progress, but it is ignored on a grand scale. Some 70 per cent of all studies in the biosciences are purportedly not reproducible. In the neurosciences figures of up to 90 per cent are in circulation. These are just estimates because very little actually finds its way into others’ cooking pots.
So scandals are rare. In July 2014, the Journal of Vibration and Control retracted 60 studies, which turned out to contain errors, at a single blow. The publisher BioMed Central has just withdrawn 43 papers. Manipulations are purported to have occurred during the review process. Of course, reviewers can miss errors. But that they do so intentionally is a new dimension. On the internet, such articles can often still be accessed despite the retraction. This is where the blog “Retraction Watch”, run by two journalists, tries to create transparency. Alongside the study itself, they publish the reasons for the retraction. Every day witnesses new entries. Some 60 per cent of all retractions are due to misconduct and not to unintentional errors.
“It is natural that researchers prefer to shine rather than admit their errors although failure can advance science.”
At this point, the auxiliary discipline of statistics comes into its own. “In Germany, there are only very few faculties and departments of statistics. At universities it is often only an optional subject, part of mathematics, for example,” says Dorothee Dzwonnek. This urgently needs changing because nowadays statistical methods are required in all sectors, she continues. Statistical dilettantism means that the results of surveys and experiments are generalised too quickly, that the number of cases used is too small and unrepresentative, or that uncomfortable data are simply ignored. Sometimes experiments are perpetuated until the desired result is achieved. The temptation is particularly strong in contract research when industrial clients have invested a lot of money. Anyone hoping for, or dependent on, follow-up funding is already in the danger zone. The number of researchers who are not dependent on state funding is increasing continually – which is not good for autonomy.
The complicity of the luxury journals
Part of the responsibility for misconduct lies with the specialist publishing houses. Researchers need to publish – only those with an impressive record are successful. In The Guardian in 2013, Nobel Laureate in medicine Randy Schekman drew a comparison with the financial crisis. “The incentives my colleagues face are not huge bonuses, but the professional rewards that accompany publication in prestigious journals – chiefly Nature, Cell and Science. These luxury journals are supposed to be the epitome of quality, publishing only the best research.” Whereby they are essentially honing their own reputation. Just as Wall Street needs to break the hold of the bonus culture, “so science must break the tyranny of the luxury journals,” he concludes. “The result will be better research that better serves science and society.”
Only those who can boast success usually have a chance of being published. Sounds logical. But it is fatal, because it engenders misconduct. “Studies that do not confirm a hypothesis will not be published. Although failure may advance science further than endorsement,” says Thomas Kühne, editor of the magazine JUnQ, an initiative of doctoral candidates at Johannes Gutenberg University Mainz. “JUnQ” is an acronym for Journal of Unsolved Questions. The magazine publishes studies, which have not produced the desired results, to prevent others from setting out on a fruitless endeavour. It is natural that researchers prefer to shine rather than admit their errors or copy what others claim to have discovered. Reproducing someone else’s findings is not seen as an achievement. There is hardly any funding to be had for it – which is why it takes such a long time before certain things come to light. Even serious offences.
Just like the case of Jan Hendrik Schön. The Austrian was a researcher in Konstanz, Germany, and the United States at the turn of this century. In 2001, the physicist published an average of one article every eight days. No one wondered about this. Years later, his scam was exposed: he had used identical measurement series for different experiments. Schön lost his job in the US, the University of Konstanz stripped him of his doctorate, and that was that. “People like Schön have several generations of doctoral students on their conscience who spent years trying to replicate their findings or develop them further,” says Kühne. Today, Schön is an engineer in a chemicals company. He was never prosecuted under criminal law.
The cheats do not have much to fear from the law
The scandal revolving around the cancer researcher Friedhelm Herrmann also turned out harmlessly for the protagonist. Around the turn of the millennium, it was proven that he had manipulated 100 out of 350 publications. The DFG wanted to make an example of the case and took it to court. But the case was dropped on the condition that he paid € 8,000. The man who was responsible for one of the biggest scandals in German research history currently has an oncological practice on Marienplatz in Munich and is still allowed to use the title of professor.
So the delinquents do not have much to fear from the law because it is difficult to prove how much harm erroneous studies really do. “In and for science it is accepted that it is a closed system that cleanses itself,” says Dorothee Dzwonnek. The loss of title and respect is the worst possible punishment. In the cases cited at the beginning, the DFG issued reprimands and withdrew the right to submit applications. “Anyone who is banned from submitting applications to the DFG over a long period of time can say goodbye to their reputation,” comments Dzwonnek.
Every year, she is presented with up to 40 cases for review. Against the backdrop of the 30,000 projects the DFG funds, this is not very much, but the DFG only promotes the best and has a strict selection procedure. This sets the bar very high. Moreover, the organisation only investigates cases that have not been reported anonymously. “We have to protect researchers from accusations of this kind, which are sometimes almost pathological.” But if someone is found guilty, their name may even be published.
Awful as it may be for an individual to be pilloried, what is worse is the loss of credibility suffered by the science system as a whole. How quickly it can decline has been demonstrated by the plagiarism cases amongst politicians in Germany. “Researchers are driven by three motives,” says Wolfgang Löwer, “and however positive the relevant vocabulary in the original German may be, it contains the very reasons for fraud and deception: curiosity (Neugier contains the word for greed), ambition (Ehrgeiz expresses a disposition to be economical with honour) and passion (Leidenschaft), which seldom spares us from suffering (leiden).”
Update: July 2018