The rate of retractions of scientific papers is rising, being up by a factor of 10 since the 1970s. But a new study shows that this is due more to fraud and misconduct, rather than simple mistakes.
The researchers from Princeton University looked at all 2,047 articles retracted in the field of biomedical and life-science research, as indexed on 3 May 2012 by the US National Library of Medicine’s PubMed service (Fang FC, Steen RG & Casadevall A 2012, “Misconduct accounts for the majority of retracted scientific publications”, Proceedings of the National Academy of Sciences, vol. 109, no. 42, pp. 17028-17033, doi: 10.1073/pnas.1212247109).
They found that only 21.3% of retractions could be attributed to error, with 67.4% being due to misconduct.
‘Misconduct’ here includes fraud or suspected fraud (43.4%), as well as duplicate publication (14.2%) and plagiarism (9.8%).
The rise in retractions has often been put down to pressure to publish leading to mistakes getting through, but the high incidence of fraud suggests that it’s more than mere carelessness.
What’s really concerning is the high rate of citations of these papers, even after they’ve been retracted. Once something is published, there’s no realistic way to control how others use it.
In the long term, these effects may even out. Assuming other researchers follow the scientific method properly, even if they start from a flawed citation their own, rigorous research will correct it. The scientific process will clean things up eventually.
But in the short term, inaccurate research can cause serious problems. An example is Dr Andrew Wakefield’s 1998 study published in The Lancet, which claimed that autism was linked to the measles, mumps and rubella vaccine. Although this was retracted in 2010 and is now considered fraud, it had a huge impact on public perception and gave ammunition to anti-vaccination campaigners.
In order to keep up with today’s relatively high number of retractions (about 1 in 10,000, compared with 1 in 100,000 in 1977), a website has been created, called Retraction Watch.
If these mistakes can be caught early enough, whether deliberate or not, then it’s hoped that the only damage will be to the reputations scientists involved.