Astronomy publications with first authors who are female receive roughly 10% fewer citations than do those with male first authors, finds a study in Nature Astronomy (N. Caplar et al. Nature Astron. 1, 0141; 2017).

In 2014, the study's lead author, Neven Caplar, a PhD student in astronomy at the Swiss Federal Institute of Technology in Zürich, attended a talk at the institute on women in science by Meg Urry, an astronomer at Yale University in New Haven, Connecticut. Inspired by Urry's presentation, Caplar and two fellow PhD students sought to confirm anecdotal evidence of gender bias in astronomy. They used data collated from the Astrophysics Data System, a comprehensive database of astronomy and physics publications maintained by the Smithsonian Astrophysical Observatory in Cambridge, Massachusetts, augmented with information from the preprint server arXiv.

The trio analysed almost 150,000 articles that were published in 5 major astronomy journals between 1950 and 2015. They found that the percentage of papers with a female first author rose from less than 5% in the 1960s to about 25% in 2015. But since 1985, astronomy publications with a male first author have received about 6% more citations than those led by a woman — a figure that Caplar and his co-authors suspected could reflect a hidden gender bias.

To determine the expected number of citations for publications led by women, the team trained an algorithm using non-gender-specific data — including the seniority of the paper's first author and other criteria — that were derived from publications with male first authors. They then used the algorithm to predict the number of citations for papers with female first authors and compared this to the actual number the papers had collected.

Caplar notes that the trio's study cannot determine the drivers of gender bias. But if women leave the field in greater numbers or earlier in their career than do men, for example, they would not be in a position to present their findings at conferences or otherwise promote their papers in ways that could boost citations.