It’s one of the worst cases of fraud ever reported in science.
Dr. Diederik Stapel, a social psychologist at the University of Tillburg in the Netherlands, has admitted to fabricating data in a significant number of studies spanning a decade of research. With over 150 publications to his name, all of Stapel’s work is now under review in an ongoing investigation led by the University of Tillburg.
Stapel’s post at the university was suspended in September, when three junior researchers reported suspicions of fraudulent data. An interim report from the investigation, released on October 31, revealed that Stapel made up data in at least 30 peer-reviewed publications. The committee expects this number to grow as the inquiry proceeds.
Already, the heavyweight journal Science, where Stapel co-authored a paper earlier this year, has issued an “Editorial Expression of Concern” to alert readers to the concerns raised over the validity of the published findings. The study, which has not yet been identified to contain fraudulent data, found that participants were more likely to engage in racial stereotyping when they were in a messy environment, compared to an ordered one.
Stapel worked in the area of social cognition, a branch of social psychology that studies how our thoughts, emotions, and behaviours are influenced by the presence of others. His research addressed topics ranging from advertising to stereotyping and discrimination.
At U of T, researchers in social psychology, whose work addresses similar questions to Stapel’s, are left troubled by the case.
“My personal reaction is a great deal of dismay that someone would perpetrate those kinds of acts, and engage in that kind of fraud in our discipline,” says Dr. Alison Chasteen, a social psychologist at U of T St. George campus whose research includes work on prejudice and stereotyping. “It’s been very disappointing and very disturbing to learn that those types of things were happening.” Dr. Michael Inzlicht, a social psychology and neuroscience researcher at UTSC, adds, “I think many psychologists — social psychologists in particular — are very shocked with what happened … and are fearful for the repercussions … [Stapel] is not a marginal researcher, someone on the fringe. He was someone very much in the mainstream, and he was very prolific. It was just a shock.” Yet the consequences of scientific misconduct reach far beyond Stapel himself. While media reports have focused on the fate of Stapel’s fraudulent publications, those in the field consider the real-life consequences for the researchers involved.
While the investigation has confirmed that Stapel acted alone in his acts of fraud, all those who collaborated with him are now also under scrutiny.
“Stapel’s students are definitely the true victims in all this,” writes Samantha Joel in an email. Joel, a graduate student in Dr. Geoff MacDonald’s lab at U of T, continues, “Their careers have been unfairly damaged, if not ruined, by the unethical actions of their advisor. Even if their PhDs aren’t revoked, many of their publications will be, which is everything when you’re trying to get a job in academia.”
According to the Tillburg committee, 14 of the 20 graduate students that Stapel supervised now hold doctoral theses containing fraudulent data. As Joel explains, “The idea of having years of work discounted through no fault of your own is unimaginable, especially when it’s the result of a person who you have come to trust as a mentor. “We’re talking about people’s lives being ruined,” Inzlicht concludes.
researchers are already talking about ways of improving the system
As for Stapel’s colleagues, the consequences are less predictable. “For the collaborators, I think they can more or less mitigate the damage, depending on how established these people are,” says Inzlicht. “But then you’ve got more junior people whose reputations are fragile to begin with — all of a sudden they’ve collaborated with this guy, and now maybe all their work might get discounted, even the work that’s not with Stapel. It’s a huge ripple effect. Especially with him, because he’s so prolific.”
While the scientific community has been working to get back on its feet following the shock of such serious misconduct, the media has stirred up a maelstrom of its own. On November 2, the New York Times published an article about Stapel following the release of the investigation’s interim report. Uncharacteristically for an article about science, the story made the front page. “One reason it’s caught the media’s attention is that in this particular case, it seems to be that Professor Stapel was very active in trying to publicize and promote his work to the media in the past,” says Chasteen. “So that sense of betrayal not only exists for those of us in the field of psychology, but probably in the media as well.”
If this is the case, then betrayal comes at a high cost. Coverage of the Stapel case has been used as a platform for broader critiques of psychological research methods, and even for scrutiny of science as a whole.
Particularly in recent years, a handful of findings have lent psychological research its fair share of controversy. An article published last year by the prominent psychologist Daryl Bem claimed to provide evidence for extrasensory perception and foreseeing the future. Bem, a well-established researcher who teaches at Cornell University, managed to get the paper published in the Journal of Personality and Social Psychology, the most prestigious journal in the field.
When the paper was released, researchers and reporters alike were shocked at Bem’s ability to publish his seemingly “unscientific” findings in the first place.
“Some people were outraged that this could happen,” says Inzlicht. “How could this be accepted? It was treated the same way as any other article. People were saying, ‘Maybe there’s something wrong with the way we accept articles. Maybe the kind of statitics we do aren’t appropriate. Maybe our very underpowered studies are contributing to this.’”
Hopefully, this will lead to greater rigour
Of course, the case of Stapel’s fraud and issues surrounding publication and peer review are very different ones. Yet, they both point to deep-set flaws in scientific research. While completely fabricating data, as Stapel did, is extremely rare, manipulating or “massaging” data is far more prevalent in science.
Inzlicht explains, “What we’re talking about right now with Stapel is outright fraud, which I think is very rare. But there are other areas where it’s a little shadier. We’re talking about what people call ‘massaging your data.’ So, looking at it one way, another way, taking out outliers, or covarying your analysis for some other peripheral measure — all to get at some sort of significant effect. “In theory this probably shouldn’t affect things that much, but it probably does. So we need to clean up our act.”
Fortunately, the Stapel case has launched discussion within the scientific community in hopes of doing just that. At U of T, researchers are already talking about ways of improving the system of disseminating research.
“We actually had a meeting devoted entirely to the notion of false positives, shortly after the Stapel case came out,” says Joel, explaining the weekly meetings held by U of T researchers in personality and social psychology. “A lot of discussion was had about how we could improve our peer review system to better prevent findings from being published that aren’t actually true. I’d say that everyone agreed that it’s a problem. The variability in opinion was more about what to do about the problem. “It was a lively discussion, to say the least.”
Chasteen adds, “Hopefully, this will lead to greater rigour in making publication decisions about papers that are submitted, greater rigour in what is expected of statistical aspects of how data is treated, and hopefully greater rigour in how people collaborate with one another — to really have a better feel for where data comes from.”
“Some people suggest that we post all of our raw data online, or in some sort of repository, so that other people can look at it,” explains Inzlicht. “They can look to see what you’ve done.” Inzlicht adds, “There’s a growing awareness, at least at U of T, that there needs to be some sort of seismic change in the way we conduct research to make sure that the effects that we discover are in fact real and replicable.”
In spite of the productive discussion among researchers, they also worry about the shadow cast on science by recent media coverage. In this case, there may very well be such a thing as bad publicity. “I do worry that the general public will start being more skeptical of the credibility of psychological findings now, expecting every researcher to be a Stapel,” Joel says. “When in truth, the vast majority of us are honest scientists, just trying to better understand how people work.”
Dr. Peter Herman, a prominent U of T social psychology researcher, echoes these thoughts. “It’s certainly too bad that the Stapel case is throwing so much negative attention at psychology — not that fraud is not a serious infraction. It’s just that it probably goes on, at a very low rate, in all academic disciplines — always has, always will. I’m skeptical that it can be entirely eliminated.”
Even so, the case exemplifies scientists’ responsibility to the public. Stapel’s misconduct not only emphasizes the importance of academic honesty for the pursuit of reliable scientific knowledge, but it also serves as a reminder that scientists rely on the public to fund them.
“All of us are beholden to federal agencies, or to our grants,” explains Inzlicht. “If now there’s another reason to suspect science, [the public will ask] ‘well now we’ve got some fraud, so why should we put money toward something like this?’
“At the end of the day, all university professors — in Canada at least — are public servants, and we’re beholden to the public. And if the public starts doubting what we do, then we’re not long for this world.”