The Washington PostDemocracy Dies in Darkness

Retractions aren’t a panacea for bad research

Placeholder while article actions load

Retractions — when an academic journal withdraws an already published study — are a big deal for researchers. Such withdrawals correct the record and remove faulty information from circulation, and they happen more often than you might realize. According to the Retraction Watch Database, over 3,100 scientific articles and studies have been retracted since 2020 in what researchers say is a reflection of tightened editorial oversight.

But retractions don’t just hurt careers and tarnish the trustworthiness of the scientific community: A study finds that they fail to reduce misinformation’s reach.

Researchers retract study that found big risks in using hydroxychloroquine to treat covid-19

In the paper, published in PNAS last month, researchers say most retractions do not happen soon enough to prevent the spread of faulty science.

The team studied nearly 3,000 retracted papers from the past decade, looking at their reach in news publications, social media and elsewhere online. When they compared the discredited papers’ reach with that of 13,500 studies that were not retracted, they found the problematic papers received more attention and were mentioned more often on news platforms than their counterparts, probably because of their compelling results.

Despite early hype, however, attention waned over time. The number of social media and news mentions upon publication were higher than papers that weren’t discredited, and the distribution of mentions suggests that most people who had been aware of the studies’ reported findings did not know about the retractions.

Some doctors spreading coronavirus misinformation are being punished

“One of the main takeaways is that retractions come too late,” Daniel Romero, a professor at the University of Michigan who co-authored the paper, said in a news release. “They remain important, but they’re not serving the purpose of reducing the amount of attention that we pay to these problematic papers because, by the time they come, the public is no longer paying much attention to the original paper.”

It’s unclear whether the increased attention that flashy research garners contributes to eventual retractions. And despite paying more attention to research that would eventually be pulled, the spotlight wasn’t always adoring. Social media users on Twitter were twice as likely to be critical of studies that were eventually removed from journals than those that stayed put.

Nonetheless, the researchers conclude, “Journals, the scientific community, and the lay public should not think of retraction as an effective tool in decreasing online attention to problematic papers.”