The past few months have been hard on scientific publishing. Several retractions have marred the reputations of some high-profile researchers as allegations of fraud were bandied about in one case and accusations of negligence in others.
Intuitively, most researchers react negatively to retractions. But some in the publishing community are happy to see that inaccuracies are being caught and removed from the record.
In August, Harvard's Marc Hauser retracted a 2002 Cognition study in which he sought to determine whether monkeys learn rules. After several months of uncertainty about the nature of the misconduct involved, documents showing that Hauser's research assistants accused him of fabricating positive results emerged. An internal investigation at Harvard concluded that Hauser was responsible for eight cases of scientific misconduct discovered in his lab, and he was asked to take a leave of absence.
In September, it was Nobel Prize-winning researcher Linda Buck's turn to disavow some of her work, retracting two papers — one published in PNAS in 2005 and one in Science in 2006 — after she found herself unable to verify the results of the studies' first author, Zhihua Zou.
Then in October, up-and-coming Harvard stem cell biologist Amy Wagers retracted a paper she co-authored with a postdoc in her lab in Nature in January; she also raised concerns over a second paper she'd co-authored with the same postdoc, which was published in Blood in 2008. The exact reasons for the retraction are not yet clear, but a Nature investigation found what looked to be duplicate flow cytometry plots in both papers.
Retraction rates are indeed on the rise. A Thomson Reuters' Web of Science analysis commissioned by the UK's Times Higher Education in 2009 shows that while the number of scientific papers published over the last 20 years has doubled, the number of retractions has increased more than 20-fold over the same time period. In 1990, just fewer than 700,000 papers had been published, and five studies were retracted. But by the end of 2008, more than 1.4 million papers were published, and 95 retractions had been issued.
The rising retraction rate could be attributed to one of two things, says Sabine Kleinert, executive editor of The Lancet and vice chair of the UK's Committee on Publication Ethics. "Either there is more fraud and more fabrication and more reason to retract papers, but it could also be that we are just getting better at pursuing allegations of such and then investigating, and actions are taken appropriately," she says. Kleinert adds she sees the rising rate of retractions as a good thing. "I'm encouraged to see a lot of retractions because I think what happened in the past is that a lot of these cases were either swept under the carpet or not dealt with appropriately," she says. "I'm quite heartened to see all these retractions. I think it's a good sign that the scientific record is being corrected."
With the spirit of collaborative research spreading through the scientific community and researchers building their work on what others have already done, it has become even more important to point out faulty research and remove it from the record. The UK committee has issued retraction guidelines for journals that state, among other things, that journal editors who come across fraudulent or faulty science have an obligation to retract the paper, regardless of whether its authors agree. "As the editor, I'm the guardian of my journal and I'm the guardian of the research record and if I have decided, following the findings of an investigation, that this paper warrants retraction — even if the authors disagree — it's my duty as an editor to retract the paper," Kleinert says.
But the rising rate of retractions doesn't mean there are a lot of bad scientists out there, she adds. It may just mean the community has become better at catching them.