A Retraction: We Gave Bad Advice
What’s behind paper retractions? (31)
by Adam Marcus and Ivan Oransky, Labtimes 06/2015
Photo: zettberlin / Source: PHOTOCASE
Better not contact the authors first.
About three years ago, we published a column in this space titled “How To Report Alleged Scientific Misconduct”. In it, we offered advice such as “contact the editor”, “resist the temptation to attack authors” and a paraphrase of the famous quotation, “Never attribute to malice what can be explained by human error”.
Much of that advice holds up, although it’s always worth revisiting as new developments appear. (Hi, PubPeer.) We’ll do that in a moment. But before we do, we want to retract one of the suggestions we made and, given how central it is, we’re going to ask Lab Times to put a prominent note on the original column directing readers to this one.
In that column, we wrote, “The first step is to contact the editor of the journal in question and ask how they prefer to receive such information, if it’s not obvious from their print or online versions. Some suggest contacting the authors first.” After a paragraph on why following stepwise procedures, as painful and frustrating as that can be, is important, we then wrote, “So, contact the authors, if the journal suggests it, or send your material to the journal directly, if that’s what they recommend.”
Three years later, after observing cover-ups, being berated by select lawyers and speaking to people intimately involved in misconduct investigations, we’ve realised that we were wrong. In keeping with some of the honest retraction notices we see, we offer apologies to the scientific community.
Here’s why: Contacting authors before anyone else knows about potential issues in their work, only serves to give unethical scientists time to hide their tracks – and let’s face it, those who are actually guilty of misconduct probably don’t have any scruples about covering up the evidence of that misconduct. That will make it much more difficult for universities and oversight agencies to investigate cases properly.
While we’d like to be able to say that we find all journal editors responsive to allegations, there are still too many who rebuff efforts to correct the literature, which means they aren’t ideal first ports of call either. That leaves the right answer: research integrity officers, or the equivalent, at the institutions where the authors in question work.
We realise that some would-be whistleblowers have little faith in institutional investigations and, where possible, outside organisations responsible for research oversight – say, the Office of Research Integrity, in the United States – are another option. Keep in mind, however, that many of these organisations are required to allow institutions to perform their own investigations first. There is, however, a paper trail, at least, and sometimes much more muscle than that.
Now, as to what has changed since our last column. At the time we wrote that original piece, PubPeer existed but it was barely on our radar, because it had only been around for a few months. We hadn’t yet seen the impact it would have, which probably first became apparent to us in May 2013, when the site played a key role in the correction of a Cell paper on human embryonic stem cell cloning. Now, of course, that impact is incontrovertible. Comments at PubPeer have led to multiple corrections and retractions. So, we now add “comment at PubPeer” to our advice.
In fact, we’ve been saying that for a few years now, whenever readers contact us about potential problems in the literature. Sure, the site’s automated notification of corresponding authors means that it has the same potential problems as going direct to authors but posting on the site also means that there’s a paper trail.
Whistleblowing and critiquing others’ work is, after all, complicated. We’re glad to have the chance to retract some of our previous advice and update it, as new evidence comes to light. That’s what self-correction is all about, right?
The authors run the blog Retraction Watch: http://retractionwatch.com
Last Changed: 22.11.2015