Are women more careful about their research than men? Or are they more risk-averse?Credit: Westend61/Getty Women are under-represented in medical research generally, but they are even more under-represented when it comes to retracted articles. A new study finds that women’s names filled just 23% of author slots in a sample of nearly 900 retracted articles published in medical journals between 2008 and 2017.
“This is a really interesting, creative and robust study,” says Curt Rice, who promotes publishing literacy at the Publishing Unlocked project in Oslo, Norway. “The article invites us to dig into issues like negotiations about authorship and the likelihood of scrutiny.” The findings were reported on 19 November in PLoS ONE1.
Why retractions data could be a powerful tool for cleaning up science Study author Paul Sebo, an internal medicine specialist and researcher at the University of Geneva, Switzerland, used an artificial intelligence tool that infers whether authors are male or female based on their first names.
He found women’s names in 16.5% of first-author slots and 12.7% of last-author slots for the retracted papers. By contrast, an earlier study that used a similar tool for predicting author gender found that women accounted for 41–45% of first authors and 26–33% of last authors in all articles from the same selection of journals across the same time frame2.
Risk and reward
Interpreting the results is fraught. The gender-prediction tools used are imperfect. They do not capture non-binary identities, for example, and can be less accurate with non-Western names than Western ones. But in a manual check of 200 names, Sebo found no mismatches.
And the findings themselves don’t come with a clear explanation. In an e-mail to Nature, Sebo wrote that his suspicion is that the disparity stems from how “women are still underrepresented in senior academic roles, lead fewer projects, and therefore may be less exposed to the kinds of responsibilities (and risks) that are more commonly associated with retractions”.