This week in Nature, Daniele Fanelli at Stanford made an interesting proposal: Set up a system of “self-retraction” that makes it crystal clear when a paper is pulled for honest error, rather than misconduct.
Fanelli, a whose work we have frequently covered, rightly notes that honest error represents a minority of retractions — around 20%. To remove any hint that a paper contains misconduct, Fanelli proposes designating self-retractions as those where all authors sign the retraction note:
To remove ambiguities, journal policies should allow authors to sign only retractions that the researchers have solicited spontaneously, because of a documentable flaw. In all other cases, retraction notes should not be signed — at least not by the authors recognized as responsible for misconduct.
Of course, authors would have to provide proof that the mistakes were due to honest error, rather than misconduct. Once that’s been validated, they should be praised for correcting the record, he notes in “Set up a ‘self-retraction’ system for honest errors:”
Self-retractions should be considered legitimate publications that scientists would treat as evidence of integrity. Self-retractions from prestigious journals would be valued more highly, because they imply that a higher sacrifice was paid for the common good. Scientists who committed misconduct would be unable to benefit.
Indeed, we’ve seen evidence to suggest that researchers who initiate their own retractions after discovering mistakes suffer no citation penalty as a result. In other words, “doing the right thing” is simply that — the right thing.
What do you think of Fanelli’s proposal? Tell us in the poll below.
[polldaddy poll=9359775]
Like Retraction Watch? Consider making a tax-deductible contribution to support our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up on our homepage for an email every time there’s a new post, or subscribe to our new daily digest. Click here to review our Comments Policy. For a sneak peek at what we’re working on, click here.
This could be tough for editors as those who had been guilty of misconduct and had been found out would be tempted to self-retract. The editors would then have the difficult problem of deciding whether there was misconduct or error. They can’t do this as they lack legal legitimacy and the means to apply due process. They would have, as with all misconduct cases, to go back to the employers.
Retraction is a must for the good of everyone. It reminds me of the quote from Mark Twain “When in doubt, tell the truth.”
And how does this deal with retraction notices that have already been issued?
Who’s to rule that it really was honest error? I don’t understand that part.
As long as the self-retraction request is the first notification that the journal gets… I wonder how hard it would be to determine when authors are self-retracting when they realize that someone (another scientist, for example) is about to publicize some strong evidence of fraud or other misconduct.
Perhaps retractions should be given a score?
Retraction (1) clear evidence of figure/data manipulation (2) evidence of plagiarism (3) retracted due to honest error?
The above is just an example of course – there are other reasons why a paper is retracted – such as a failure of one set of authors to correctly attribute other researcher(s) for example (or a paper was published without the other authors on the paper knowing about it).
A score would allow researchers in the field to determine the reason of a retraction quickly when looking for relevant papers.
Just a thought.
I think the message remains the same in either case: the paper should be considered “withdrawn” from the scientific literature. Assigning “guilt” is an entirely different process, much less objective, especially when people disagree on whether it was honest error or not, so that question can be kept separate from the decision to retract. In any case, it should suffice if the retraction notice explains how the retraction came about, that is whether it was initiated by (some of) the authors or the editor (perhaps following a comment from some anonymous third party). That can already provide the same information.
I guess this is another aspect of making retraction notices transparent…?
I agree with Richard Smith. It is very difficult for an author to “prove that he or she made an honest error” — they can only claim to editor that they made an honest error and explain how that happened, and show documentation if there is any. If the editor finds that explanation to be questionable, perhaps a misrepresentation of the facts, then the editor must ask the institution. For an investigatin
A retraction notification should make it clear to readers whether there was misconduct or honest mistakes, so I do not understand why we would want/need to differentiate them. I agree with Dave Langers that the paper should be considered withdrawn from the literature in either case.