On Retractions—and a Retraction Watcher

When a journal article is published, it normally stays published.

Sometimes, however, journal articles are retracted. In other words, they are declared invalid because they are found to have overwhelming flaws. Although published copies remain, these articles are no longer considered valid publications.

Retractions occur for various reasons. Sometimes the article was found to contain fabricated or falsified data. Sometimes the article was found to be plagiarized.

Not all retractions, though, reflect misconduct. Sometimes researchers discover major flaws in their published research. For instance, they may see that an analysis was wrong, thus invalidating their conclusions. In such cases, honest researchers retract their papers.

Recently I attended a talk by an expert on retractions. The talk was by Ivan Oransky, who co-founded and co-runs the blog Retraction Watch. Oransky (see photo) was being honored by a chapter of the American Medical Writers Association.

Oransky noted that the retraction rate has been rising. Why the increase? Maybe, Oransky said, fraud is becoming more common. And maybe, he said, problems are being discovered more, for example because of plagiarism-detection software.

Articles sometimes are widely cited after retraction, Oransky noted. Avoiding this problem might become easier, though, because an organization established by Oransky and a colleague recently received a grant to create a database of retractions.

Oransky ended his talk with an encouraging finding: According to a recent study, researchers who do the right thing by retracting their own papers do not diminish and might enhance the number of citations of their other work. "Scientists reward authors who report their own errors," he said.

Until the next post—

Barbara