Why Retraction Notices Matter

What’s behind paper retractions? (13)
by Adam Marcus and Ivan Oransky, Labtimes 06/2012

Photo: zettberlin / Source: PHOTOCASE

Two recent studies find misconduct is more common than scientists like to think.

By the time this column goes to print, Retraction Watch will have tallied its four-millionth page view. Whether that’s a lot or a little is less important than what’s happened since our two-millionth view: quite a bit. Our understanding of retractions has undergone a fairly radical shift since we first began the blog, and, indeed, in the past month or so.

What has changed? Thanks to a pair of recent studies, we now know that, contrary to previous reports, the majority of retractions in the life sciences – two-thirds, to be precise – result from misconduct, not honest error or other innocent causes. Although most misconduct does not rise to the level of fraud, the new research certainly undermines the assumption that an interest in retractions is marginally better than prudence.

More info, better stories

As far as Retraction Watch is concerned, we have, since our inception, pushed for journals to be more transparent with their retraction statements. That argument might have seemed self-serving. After all, the more we know about retractions, the better the stories we can tell. But many agreed that opening the books gave readers of scientific journals important, and sometimes critically important, information about their fields.

And these new studies – one in PNAS and the other a National Bureau of Economic Research working paper – demonstrate why. When the authors, knowing that so many retraction notices are wordsmithed to uselessness, looked at Retraction Watch posts, Office of Research Integrity findings and other secondary sources, they found that many retractions seemingly due to error or unclear reasons were actually caused by misconduct. (A third study in PLoS ONE, while very helpful in gathering retractions all the way back to the 1920s, is less useful for determining how common misconduct is because it took notices at face value).

For us, our contributions to a better understanding of the reasons for retractions are far more of a validation than millions of page views. There are also far more conversations taking place about correcting the scientific record than there were before. Even Elsevier, once wary of our goals and motives, said of Retraction Watch, “Scholarly publishing is better for it.”

Painstaking efforts to set the record straight

It might be tempting, in a strange way, to say that we’ve nudged ourselves toward irrelevance. We still find plenty of notices that say nothing, almost as if challenging us to reveal the misconduct investigations that gave birth to them. (Don’t worry, we love those challenges). But we also see painstaking efforts to set the record straight with illuminating notices. Sometimes, editors even cite Retraction Watch, as either a threat or a guide to best practices. That flatters us but we wish it was unnecessary.

Over the past two years, we have seen a shift in the willingness of journals to be open about the retraction notices they print. In light of the new data, we hope they’ll continue to move in that direction.

And if Nathan Georgette, a senior at Harvard University, is any indication at all about the next generation of researchers, transparency is trending positive. Georgette – who published three papers in high school – realised he’d made an error in one of them once he took a more advanced math class in college. So he went ahead and retracted it, apologising for his inadvertent mistake. The notice is clear and he was more than happy to answer follow-up questions from us.

Doing the right thing

Compare his behaviour with that of many senior scientists, who ignore legitimate criticism and find ways to not retract. Here was a case in which no one – not peer reviewers, not readers – even noticed any errors and Georgette retracted anyway, because it was the right thing to do. Even if he’s is an outlier, there’s no denying his open approach is worth emulating.

(The authors run the blog Retraction Watch: http://retractionwatch.com)

Last Changed: 07.02.2013