Women are inadequately represented as peer reviewers, journal editors and last authors of studies, according to an analysis of manuscript submissions to an influential biomedical journal.
The study looked at all submissions made to the open-access title eLife from its launch in 2012 to 2017 — nearly 24,000 in total. It found that women worldwide, and researchers outside North America and Europe, were less likely to be peer reviewers, editors and last authors. The paper — which hasn’t itself yet been peer-reviewed — was posted on the preprint server bioRxiv1 on 29 August.
About 7,000 of the submitted studies went through the full submission process (at eLife, authors make a ‘pre-submission query’ before being invited by the journal to send a full paper — a relatively uncommon practice among journals). In all, the analysis covered the activity of about 7,000 referees, 890 reviewing editors and 57 senior editors.
The researchers found that women make up only around 20% of peer reviewers, and around one in four reviewing editors (see ‘Peer-review patterns’). Most reviewing editors and reviewers were in the United States — 62% and 56%, respectively — followed by the United Kingdom and Germany in second and third place. Less than 2% of peer reviewers were in developing nations — all in China, India or South Africa.
Of the full submissions, 1,549 (22%) had a female last author — a position that indicates seniority — and 5,127 had a male last author. About 53% of manuscripts with male last authors were accepted, versus around 50% of those with female last authors.
Fifty-seven per cent of fully submitted papers with a male last author were accepted when the review panel was all male (see ‘Peer-review patterns’), whereas mixed-gender teams accepted 51% of male-last-author papers. And submissions that had been edited or reviewed by someone in the same country as the corresponding author were more likely to be accepted than those with a country mismatch.
The trends are likely to be a result of implicit biases, says study co-author Cassidy Sugimoto, an information scientist at Indiana University Bloomington. The study did not seek to reveal how the disparities came about, say the authors. But because the gender make-up of senior authors and gatekeepers closely matches disparities found more broadly in academia, there is no evidence that eLife is making such disparities worse.
Body of evidence
The research was prompted by eLife, which approached Sugimoto and her colleagues with the data; two study authors are eLife employees. The journal’s reviewing process is unorthodox in that referees know each other’s identities, which allows them to discuss any differences of opinion on manuscripts.
The study is robust, says Jevin West, an information scientist at the University of Washington in Seattle. And it is concerning that women and authors in developing countries seem to be marginalized in peer review, he says. “It’s very important that we have diverse voices represented and that those voices are treated equitably.”
The results echo previous findings about peer review. This month, a global survey by Publons — a site that allows academics to record their peer-review activity — found that researchers in developing countries are under-represented as reviewers, yet more likely to accept review requests and complete reviews faster.
And last year, an analysis2 of American Geophysical Union (AGU) journals found that women are invited to review less often than expected, but that the editors’ gender has no influence on acceptance rates. Brooks Hanson, executive vice-president of science at the AGU in Washington DC, says the latest research chimes with previous findings that female editors engage more female researchers as reviewers.
Sugimoto says that journal policies should aim to ensure diversity on review panels, for example, by inviting a greater proportion of women and researchers in developing nations to do reviews. “This is one of the simplest policy changes we can make,” she says, “without high risks, and potentially high benefits.”
Andrew Collings, eLife’s executive editor and a study co-author who is based in Cambridge, UK, says that the team is communicating its results to the editorial board, so that editors can consider the findings as they assess submissions and select reviewers. “We are particularly keen to see editors using diverse groups of reviewers whenever possible.”
To weed out the effect of implicit biases on acceptance rates, it is tempting to see blinding as a solution, West says. But, he adds, double-blind peer review — in which neither authors nor reviewers know each others’ identities — often works poorly, because some fields are so small that reviewers can easily guess who wrote a paper.
Sugimoto says that more data are needed to determine the effectiveness of techniques such as blinding or open peer review, in which reviews are published online and authors and reviewers may know one another’s identities.
She hopes that more journals and publishers will release data on peer review for analysis. “Then, we can inform it with evidence rather than with anecdote.”
Nature 561, 295-296 (2018)