Our comment threads lit up today with news that the interim report on the misconduct investigation into Diederik Stapel has arrived — and what it says ain’t pretty.
Stapel, a Dutch social psychologist, was suspended in September from his post at Tilburg University over suspicions that he had used fabricated data in some of his published articles. According to the report, the misconduct is sweeping, involving at least 30 papers, and possibly as many as 150.
At least as troubling, it seems that 14 doctoral dissertations also might be impugned by Stapel’s bogus results.
Of course, there will likely be retractions. We’ll continue to follow the case.
Hat tip: Thanks to many posters, and Willem van Schaik, for bringing this to our attention.
This is big, and embarrassing for many. Three quotes from the report that I find most striking (my translation):
p.14: “It is important for a PhD-student or research master student to gain experience with the entire research process, including gathering and analyzing the data, especially for one’s own research. A number of students of Mr. Stapel have therefore never experienced this process themselves.” (About how 14 out of 21 PhD-students could have worked with data that were made up without them being aware of it.)
p.16: “Only verification [of hypotheses] counted. Anyone with any research experience, especially in this sector [i.e. social psychology], knows however that most starting hypotheses prove false. And even if they turn out to be valid, the effect often disappears during replication.” (A sobering but realistic assessment of the scientific level of the field, in my opinion.)
p.25: “I find it important to stress that the errors that I made did not arise out of self-interest. I do not recognize myself in the image that is drawn of a man that tried to exploit young researchers for his own benefit.” (Stapel, briefly reacting on the statements in the report.)
Other bits:
– The commision (that wrote the interim report) says that Stapel’s PhD cannot be reprived on the basis of data manipulation (as the data is not available anymore), but recommends an investigation (by the University of Amsterdam) into the possibility of taking his PhD away from him on the basis of ‘exceptional scientific misconduct’.
– Stapel states in tha report’s attachment that he is ‘dismayed and ashamed’ as a result of the report, that he is unable to reply to the accusations at the moment due to his prsent condition, but that he plans to issue a statement in reply to the report next monday.
Two more additions:
– The report was issued by Tilburg University (http://www.tilburguniversity.edu/). Groningen University is performing an independent investigation, but did cooperate with the Tilburg group and added comments/clarificatuon at some points in the report. (as did their collegues from Amsterdam). You might want to switch logos….
– While those 14 dissertations might also be impugned, the commision stresses that the authors are not to blame. In none of the cases investigated thus far,there should be action taken against the authors (i.e. they can all keep their PhD’s). Groningen University adds the suggestion that the authors of these dissertations should perhaps receive an official declaration of this.
Oh, and sorry for the typos and/or poor grammar. Usually I quickly proofread my English ramblings…
Recently, and for diverse reasons, I’ve looked back at the history of Marc Hauser, reviewed the extended American Anthropology Association saga, re-read parts of Pinker’s “Blank Slate,” pondered the peculiar collapse of Stephen Gould’s case against 19th century brain research, and now followed the Stapel story. The question, not to put too fine a point on it, is this: what do we do if an entire field of science is so infected with fraud and agenda-driven literature that it simply isn’t trustworthy any more?
Maybe that’s the case with social psychology. Maybe it’s not. But it’s a problem which is simply no longer excusable as some dilute Poisson-distributed behavior by a tiny group of three-sigma deviants. The Stapel matter isn’t a scandal. It’s one more symptom of a real level of systemic rot in this particular discipline. It needs to be addressed as such.
Re: “what do we do if an entire field of science is so infected with fraud and agenda-driven literature that it simply isn’t trustworthy any more?”
We need to work with the strengths of science, that science is capable of self-correction. We need to spend time and resources to replicate key studies; replicate blindly, if at all possible. We need to acknowledge and discuss emerging problems openly, to facilitate self-correction.
The Dutch interim report details why the typical self-correction mechanisms don’t work. Social psych data is almost always in the form of anonymous surveys. Since subjects usually won’t respond honestly without strict confidentiality, the data is deliberately made untraceable. That’s only good science and good ethics, right? One also needs lots of personal information to control for, or identify, factors like drug use, family issues, or sexual orientation and history. So, as Stapel explained to doubters, it was important to keep the total trust of the sample populations and restrict actual survey work to the one trusted, senior person, Stapel himself.
In addition, take a look at the results reported or attacked in the cases I mentioned. Stapel, for example, had an absolute genius for telling his academic audience what they wanted to hear. Stapel’s typical paper seems to have reported some unpleasant behavior (racism, sexism, hypocrisy) and tied it to vaguely defined authority groups or their presumed reactionary attitudes. As an academic, are you seriously going to go out on a limb to question these results? Based on what data? As the Dutch report pointed out, the usual journal doesn’t even report social psych methods in any detail, much less present the data. As a potential critic, are you even going to risk exposing your possible ignorance of, or lack of sophistication in, the undoubtedly subtle statistical methods used by the esteemed professor?
Social psych has relatively unique characteristics which, ironically, make it a magnet for academic sociopaths. The usual methods of scientific self-correction aren’t totally disabled, but they work poorly and very slowly in this discipline. The Dutch administrators who charged the investigators were right to be thoroughly alarmed about the state of the field.
Wow. Hats off to Tilburg and Groningen…They released the report in approximately a month or so.
For people who understand Dutch, or use Google Translate
the different ways how Stapel conducted the fraud is breath taking
http://www.tilburguniversity.edu/nl/nieuws-en-agenda/commissie-levelt/interim-rapport.pdf
I feel bad for the PhD students who wasted years of their life working on bogus projects!
Stapel is the “scientist” who claimed that eating meat makes you selfish and antisocial. As a vegetarian, I’m embarrassed by this guy. Social psychology and evolutionary biology are full of “studies” that claim that people who are not like the average evolutionary psychologist have mental deficiencies (remember when they were saying that liberals are smarter than conservatives)? This sort of “research” seems like dangerous territory. It seems like the only people who are interested in studying such questions are likely to be crazy, and they also are probably looking for personal validation.
This case is breathtaking enough without exaggerations, one would think. Still, incorrect numbers and statements are flying around. I commend Adam Marcus for the careful wording: “it seems that 14 doctoral dissertations also might be impugned by Stapel’s bogus results.” Mainstream media word it more spectacularly: “14 of 21 of the theses published by Stapel’s students were affected by the tainted data” (e.g., http://www.washingtonpost.com/blogs/achenblog/post/diederik-stapel-the-lying-dutchman/2011/11/01/gIQA86XOdM_blog.html who get the prize for the best headline). Wikipedia has followed suit with “He generated fake data for 14 of the 21 dissertations he supervised.” Unfortunately, this is wrong.
So what are the facts? Please count along with me on page 6 (last para) and 7 of the official interim report (http://www.tilburguniversity.edu/nl/nieuws-en-agenda/commissie-levelt/interim-rapport.pdf). Even without any Dutch language skills, you can count 19 names that are listed, not 21. Of those, 7 are cleared of all suspicions (to be precise: one in the last para on p. 6, six more in first sentence of the last para. on p. 7). 19-7=12, not 14.
Someone seems to have made up two entire PhD’s (even Stapel wasn’t that good 😉 To quote Adam: “there will likely be retractions”!
So let’s look more closely at what is said about those 12. The degree of suspicion varies quite a lot upon close reading.
First off, the report states none of them are culpable: They did not make up data themselves, nor can they be reproached for not having found out what Stapel was up to. They were all deceived.
So how badly are the 12 victims affected?
-For 3, there is a factual statement that “one or more chapters” contain fake data (these dissertations contain 4, 4, and 5 chapters with data respectively. Each chapter contains 1-8 studies. I haven’t calculated this, but I the median might be 3)
-For 2, it says there are signs that fake data may have been included in “minor parts” (the Dutch phrase used here is “ondergeschikte delen” which conveys that they are parts that are of lesser importance. I can only speculate about the reasons for using this phrase: they could refer to pilot data, perhaps?).
-For the remaining 7, there are “doubts.” It says that in some cases data were given to PhD students by Stapel, in other cases the data had been “in some stages of the analysis” in the hands of Stapel.
So I believe that Adam’s statement should be reworded: 12 dissertations might be impugned. Of these, some are badly affected, some less so, some are unclear. Let’s wait for the results before jumping to conclusions.
Also check out the NYT Online report Nov 2 about this mess; says social psychology is riddled with these problems. The only answer is careful, repeated attempts at replication of the most controversial findings first, then the rest of them…
Google translate–
You read the personal reaction of Professor Stack below.
October 31, 2011
In recent weeks I’ve been thinking whether I should respond and, if so, what I have to say. It’s hard to find the right words. The committee has spoken. And now I have and I want to say something, how impossible it is right to say.
I have failed as a scientist, as a researcher. I’ve updated research and studies feigned. Not once but several times, and not equally, but over a longer time. I realize that through this conduct my immediate colleagues in dismay and anger and I left my field, social psychology, have put in a bad light. I feel ashamed for it and I have great regret.
Science is about people, it’s teamwork. I have in recent years enjoyed working with many talented, highly motivated colleagues. I committed them to emphasize that I never have been aware of my inappropriate behavior. I offer my colleagues, my doctoral students and the entire academic community, my sincere apologies. I am aware of the suffering and the grief I have caused them.
Social psychology is a big, interesting, beautiful and strong discipline, provides unique insights into human behavior and therefore still a lot of attention. I made the mistake that I see the truth in my hand I want to put the world and just wanted to make something more beautiful than it is. I have used improper means to the results attractive. In modern science is the ambition level is high and the competition for scarce resources is enormous. In recent years, these pressures become too much for me. I have the pressure to score, to publish, the pressure to always be better, not have faced. I wanted too much too soon. In a system where there is little control, where people often work alone, I am wrong repulsed. I would stress that the mistakes I’ve made, not born of self-interest.
I realize that there are many questions. My current state constitution does not permit me to answer. I will still have to dig deep to find out why all this happened, what has moved me to do so. I need help I have now also been given.
Here I want to leave right now.
Diederik A. Stack
I am confused. I had thought that this was about Diederick Stapel. You use the name “Stack”. Do you need to retract your comment?
“Stack” is how Google translates “Stapel” from Dutch.
Haha, yeah. I saw the translation and left it in there on purpose 😉
”I committed them to emphasize that I never have been aware of my inappropriate behavior.” is not well translated by Google Translate, should be something like this: ”I want to emphasize that they
[my colleagues] were never made aware of my inapproriate behaviour”
What about a URL link (to the original Dutch)?
http://www.scienceguide.nl/201110/stapel-reageert-op-fraude-rapport.aspx
@M.A.D. Dank u wel voor de pagina.
Nature nicely sums it up by saying that there is a “Massive fraud at Dutch universities” . Thanks for the generalization there, Nature editor. Read the article. Also watch for the cheap shot directed against Science, which published his work.
http://www.nature.com/news/2011/111101/full/479015a.html
The case is travelling worldwide, see links below:
http://www.reciencia.com/?p=993
http://cienciabrasil.blogspot.com/2011/11/mais-um-mega-fraudador-da-ciencia-e.html
I wish every fraud case would get this attention and coverage. Loved that his speech was posted here.
In Brazil the massiveness of fraud produced by one person is only paralleled by a chemist Dr Airioldi (also featuring in Retracton Watch) yet Brazilian academia hushed the scandal. Many suspect he defrauded over 50 papers with Photoshop and sheer invention. In terms of visibility, a Brazilian case comparable with this one was with Drs. Gomes & Zuben, also commented in Retraction Watch; the authors made wonderful public statements claiming that no harm was intended, and then threatening anyone who called them ‘plagiarists’ or ‘bandits’.
This is a good development of modern science, and Retraction Watch is playing a central role.
My comment, which comes from my general reading of this case is: did any of the people who now have likely useless (perhaps to be retracted) PhDs based on made-up data ever bring any concerns to the attention of the authorities? I do wonder if these people were simply ignored.
I just read that yesterday, Diederik Stapel has voluntarily given back his Ph.D. certificate to the University of Amsterdam. The UvA press notice (in Dutch) can be found at :
http://www.uva.nl/actueel/nieuws/nieuws.cfm/F684F930-9ACC-4591-B5A6EAD9F74F5F10
The press notice reports that Mr. Stapel has declared in writing that he gives back his degree because he considers the behaviour he engaged in in the last couple of years unfitting for someone holding a Ph.D.