There are a number of barriers to correcting or retracting scientific papers, the University of Alabama at Birmingham nutrition researcher David Allison and his colleagues write in Nature.
While Allison had success getting one flawed paper retracted — a mathematical model it relied on overestimated the effect the researchers observed 10-fold — he and his colleagues write that that was an exception. They uncovered more than two dozen papers in their field with substantial errors and tried to get them addressed before it ate up too much of their time and they had to stop.
"We learned that post-publication peer review is not consistent, smooth, or rapid," Allison and his colleagues write.
They ran into the same six problems repeatedly: there's no standard way to request raw data; informal expressions of concern such as comments on papers aren't prominent; it's often unclear where to send expressions of concern; journal editors are unwilling or unable to take action; journals that acknowledged errors were still reluctant to retract; and some journals charge researchers to correct others' errors.
One publisher, they add, said it would charge $10,000 to whoever initiated the withdrawal of a paper.
At Stat News, Retraction Watch's Ivan Oransky and Adam Marcus add another stumbling block: lawyers. Nature, they note, has said in an editorial that journals contemplating retractions sometimes are threatened with lawsuits.
Allison and his colleagues argue that guidelines, like the ones that have been developed for paper submissions and peer review, also need to be drawn up for post-publication peer review.
"Robust science needs robust corrections," they write. "It is time to make the process less onerous."