The modern scholarly publication system serves as the primary means of communicating scientific results, typically through peer-reviewed articles. While the peer-review process attempts to keep incorrect, plagiarized and fraudulent work out of scientific journals, it doesn’t catch everything. As a result, journals very occasionally need to un-publish an article, a process known as retracting the article or a retraction. The journal is basically taking back everything that was said in the article.
Articles can be retracted for a variety of reasons. Some articles are retracted because others can’t reproduce the same results, even though the researchers acted in an honest and ethical manner. Other articles are retracted due to misconduct on the part of the authors. Plagiarizing another article (including another article published by the same authors), submitting deliberately fraudulent results, or ethical violations are all grounds for retraction.
An article can be retracted by various people. One (or more) of the authors can request that an article be retracted. Sometimes comments from the scientific community can lead to a retraction. An editor or publisher can also start the retraction process.
What happens next varies widely from journal to journal, based on publisher policies (or lack thereof) or editorial practices.
Once the decision is made to retract an article, the journal typically posts a retraction notice on its website in place of the article. The information within the retraction notice varies widely. Some simply state, “This article has been retracted.” Others provide details about why an article was retracted. Some publishers leave a copy of the original article on their website, with large notices that the item has been retracted.
The excellent blog Retraction Watch routinely posts information about retraction notices, and advocates for more transparent retraction notices. Like many aspects of our modern life, journal publishing has become more litigious. Authors of articles recommended for retraction have threatened legal action against publishers, and the editors of Nature recently cited legal threats as one reason for their delayed retractions and vague retraction notices.
Somewhat counter-intuitively, the most prestigious journals tend to have the most retractions (we are still only talking about a handful of articles each year). An analysis by Morrison (2011) suggested that journals with higher impact factors are more likely to retract articles than those with lower impact factors. These high-impact journals tend to publish articles that have the most novel conclusions or that advance our knowledge of a subject by a lot. As a result, these journals sometimes publish articles in which an innocent error led to a more definitive result than the data really suggest. Alternatively, some researchers feel pressure to publish in high impact journals, leading them to fabricate results.
Sometimes an article is retracted quickly, before it is cited in the scholarly literature. In other cases, an article can be cited many times before the retraction. There is no formal mechanism to notify authors of papers that cite a retracted paper, and the importance of the retracted paper to the citing paper varies widely. In some cases, retracted articles continue to be cited in the scholarly literature if authors fail to go back and read the articles they cite, relying on the citations in another paper.
Overall, the rate of retracted articles has increased recently (Steen et. al, 2013), possibly due to changes in the behavior of both authors and institutions. However, Steen et. al, 2013 also suggest that the scholarly publishing system is getting better at finding erroneous or fraudulent work and retracting those articles.
Although the system is far from perfect, the retraction process is one of the ways in which the scientific literature corrects itself over time.
Fang, F. C., & Casadevall, A. (2011). Retracted science and the retraction index. Infection and Immunity, 79(10), 3855–9. doi:10.1128/IAI.05661-11
Retraction challenges. (2014). Nature, 514(7520), 5. doi:10.1038/514005a
Steen, R. G., Casadevall, A., & Fang, F. C. (2013). Why has the number of scientific retractions increased? PloS One, 8(7), e68397. doi:10.1371/journal.pone.0068397