On science blogs this week: Fraud again
UNTRUSTWORTHY SOCIAL SCIENCE PAPERS, PART II: THE LYING DUTCHMAN. Hot on the heels of last week's discomfiting report on untrustworthy data analysis in social science research comes this full-blown, much higher-profile social science data scandal.
This time it's outright fraud, about a (formerly) respected Dutch social psychologist, name of Diederik Stapel, who seems to have saved himself a ton of the time and trouble it takes to find subjects and actually study them. Instead, he simply made up his data. For years. In dozens of papers (perhaps as many as 150!) While employing a series of tractable grad students who analyzed the data he gave them, apparently not knowing it was fictional. Fourteen of their degrees are now in jeopardy.
Diederik Stapel. Why is this man smiling?
Last week I cited evolutionary psychologist Robert Kurzban's dismay at a paper showing that much of the social science literature may not be trustworthy, partly because analytical practices in the field encourage false-positive results. Kurzban has written about scientific fraud before, including the Stapel case. In September he had this to say:
...as far as I know, commentators on these scandals do not seem to think that what these individuals did reflects in some way on the entire discipline in which they work. I’ll be surprised if the Stapel matter leads anyone to say, out loud, for instance, that somehow his actions should lead us to believe that social psychology is a hopeless wasteland of pseudoscience, replete with superficial or vacuous theory and parlor tricks that pass for scientific experiments.
Ummmmm... That is in fact pretty close to what the paper Kurzban discussed last week was saying. However:
Razib Khan at Gene Expression agrees with Kurzban, not only on this point, but also last week's about false results from more subtle causes.
Here’s the controversial thing I’m going to say: instances of flamboyant fraud are probably far less of an issue in a discipline like social psychology than more subtle biases and systematic incentives...‘Sexy’ social psychological research is routinely blasted by the press, and devoured by the public. Incentives matter, and Diederik Stapel is just the reductio ad absurdum of scientists who respond to the repeated 15 minutes of fame which the press provides. Most of the time it’s not so blatant, nor frankly so malicious in consequence. I doubt that there are many Diederik Stapels in the field, else he wouldn’t have gotten so far with bluster and bluff.
Not clear to me why he thinks that's controversial, especially after the paper about false positives.
The Stapel story has been around for a while, but commanded blogging attention this week because one of the papers in question, a much blogged-about tale purporting to show how a messy environment influences ethnic prejudices, was published earlier this year by Science, and Science was irritated enough to issue an "expression of concern".
At Retraction Watch, Adam Marcus has been keeping an eye on the Stapel story for some time. He wrote about Stapel's suspension by Tilburg University, when the University of Groningen launched its investigation, about the Dutch interim report on Stapel's misconduct, and about the Science "expression of concern".
At the Chronicle of Higher Education Percolator blog, Tom Bartlett gives us details on how Stapel fooled so many for so long. At her sprightly blog HEALTH'Sass, health reporter Star Lawrence was exasperated by this and other research malfeasance:
You are wasting my time, people, and dragging down big salaries while doing it. I don’t like it.
In 1999, Stapel and some colleagues published a (possibly fabricated) study on a Dutch plagiarism case. Marcus marvels at his gall, and at the gall of other fraudsters who have gone public in the same indirect way.
But why is that surprising? These folks are inveterate, devout, risk-takers. You don't stumble into a high-wire act like this by accident. I suppose it's possible that these cloaked self-revelations are unconscious. But isn't it even more likely that they are examples of calculated nose-thumbing, a case of épater le bourgeois? Only in Dutch?
By the by, I can't pretend that I thought up that brilliant play on words in the hed, The Lying Dutchman. Such a claim would be, uh, fraudulent. It's lifted from a post on this tale by Joel Achenbach, master of the Washington Post Achenblog — and master punster.
WHATEVER HAPPENED TO THE NATIONAL PRACTITIONER DATA BANK? The National Practitioner Data Bank, a valuable asset particularly to journalists because it contains information about discipline of physicians and malpractice awards, is once more freely available, thanks to Investigative Reporters and Editors, which has put it back online.
The National Practitioner Data Bank is a public government database that had been online for years until the Obama administration shut off access in September. At the New York Times blog Prescriptions, Duff Wilson provides useful background on the database and response to its removal.
...the public version of the database does not identify physicians by name or address, but it does provide other useful information about hospital sanctions, malpractice payouts and state disciplinary actions against every doctor in the country.
A series of posts at AHCJ has tracked the controversy. Jeff Porter, for example, recounts how the database resurfaced and how AHCJ, IRE, and the Society of Professional Journalists have protested the government removal of the database from public view. Most recently, Len Bruzzese reports that the U.S. Health Resources and Services Administration took down the public version of the National Practitioner Data Bank at the urging of a Kansas neurosurgeon with a long history of malpractice payouts. The revelations were contained in records released by US Senator Charles Grassley.
The Health Care Blog published an open letter to government officials from several medical academics. They complained:
By pulling down the [public database] at the request of a single physician with 16 prior malpractice claims, the HRSA took a large step in the direction of the “bad old days” when secrecy prevailed and providers’ interests took precedence over patients’ safety and well-being. Fortunately, we no longer live in those times.
I admire their ability to take such a rosy view of our times. If only.