Science is still mostly a boys' club.
Fewer than a quarter of all science, technology, engineering, and mathematics (STEM) jobs in the US are held by women. To figure out what exactly is going on and how to remedy the disparity, researchers have been hunting for all the places in STEM fields where gender biases, both explicit and implicit, are prevalent. These biases have been found in hiring decisions, grants awarded, award nominations, and more.
One such disparity shows up in a more subtle, but still influential, place, according to new research published Wednesday in the journal Nature. Cross analysis of membership and authorship data from the American Geophysical Union (AGU) and the organization's journals reveals a gender disparity among peer reviewers. And identifying this gap and understanding where it's coming from could help guide the AGU, and science generally, toward greater gender parity.
While 28 percent of AGU members were female, and 27 percent of the first authors listed on papers published in AGU journals from 2012 to 2015 were female, just 20 percent of all peer reviewers were women. Men conducted 2.1 reviews per person on average, while women conducted 1.8 reviews.
"The differences may seem slight," Brooks Hanson, the director of publications at the AGU and an author of the Nature commentary on the AGU data, tells The Christian Science Monitor. "But that builds up over time."
The often unpaid task of critiquing a colleague's paper may seem secondary to conducting original research, but it's an important component of a scientist's career development, says study author Jory Lerback, a graduate student at the University of Utah who was previously a data analyst at the AGU. It can help scientists engage with and contribute to cutting-edge research in fields outside of their own work, develop their own writing and critical-thinking skills, build reputations for themselves, and network with potential collaborators.
And, Ms. Lerback adds, peer review is largely about ensuring that a scientific paper considers all the possible angles, issues, and viewpoints of a topic, "so having a diverse pool of reviewers is very important for science generally."
So where are the female scientists getting lost in this process?
Lerback and Dr. Hanson's analysis found that many women were being missed as peer reviewers because authors simply did not think of suggesting them, or editors did not think to ask them.
When a scientist submits a paper to one of the AGU journals, Hanson explains, they are asked to recommend a handful of other experts in the field that might be able to conduct a peer review. The editors of the journal then consider that list and their own knowledge of experts on the topic, and then invite a couple of scientists to be peer reviewers.
The disparity wasn't just coming from male first authors and male editors. Female authors and editors weren't thinking of women frequently either.
Lerback and Hanson found that female first authors suggested female reviewers 21 percent of the time, while men suggested women 15 percent of the time. Female editors invited female reviewers 22 percent of the time, while male editors invited female reviewers 17 percent of the time. All of those proportions were smaller than the overall pool of female AGU members (at 28 percent of the whole population).
One might think that these numbers are a holdover from historic trends, since women have historically been less-well represented in science and therefore fewer would be longtime experts in a field, Hanson says. So Lerback and Hanson broke it down by age (as a proxy for experience). And they found that in each age cohort, women are consistently underrepresented as peer reviewers.
A possible solution: Gaining recognition
One solution that Hanson says the AGU is already working to employ is to have more diverse editorial boards, as female editors were found to invite female scientists to review papers more often. Another could be to help female scientists become well-known in their field in other ways, perhaps with awards, as peer reviewers are called upon based on recognition, visibility, and connections within a network of scientists.
Female scientists are receiving awards, but they're not often for the women's research. A study across 18 STEM disciplinary societies conducted by the Association for Women in Science (AWIS) found that men receive a higher proportion of scholarly and research awards than women (about eight times as many), while women receive more teaching and service awards, relative to their representation in a given field.
Resolving this disparity could help raise female researchers' visibility, reducing one cause of the gender disparity in peer reviewer selection, suggests Heather Metcalf, the director of research and analysis at the AWIS, who was not involved in the AGU study, in a phone interview with the Monitor.
Dr. Metcalf and her colleagues at the AWIS have a few ideas about how to do this, and have already conducted workshops with organizations, with successful outcomes. The language used in nomination materials, award titles, and other materials could be gendered, she suggests, which could have an effect on who thinks they might be qualified. The make-up of nominee pools and awards committees could also have an effect.
At the root of it all, Metcalf and others suggest, may be unconscious biases that are pervasive throughout. When authors and editors consider potential peer reviewers, they are thinking of experts in the field. So the disparity, Lerback suggests, could be due to implicit biases in what an expert looks like – and that image might not be of a woman.
Previous research found that male researchers were more likely to receive excellent reference letters for postdoctoral positions than female scientists, which could indicate a bias viewing women as innately less capable of being scientists than men. Similarly, another study found that many researchers consider a "spark of brilliance," or an innate high-level of intelligence commonly attributed to men, necessary to succeed in their respective fields.
To tackle these implicit biases, Anna Kaatz, director of computational sciences at the Center for Women's Health Research at the University of Wisconsin-Madison, suggests making "the unconscious conscious."
But, she says in a phone interview with the Monitor, simply raising awareness is not enough and can actually make things worse. "If you raise awareness about an issue, and you don't teach people what to do about it, it essentially paralyzes them … people quit talking to each other, they get nervous, they don't know what to do."
That's why people like Dr. Kaatz and organizations like the AWIS are building workshops for organizations, editorial boards, institutions and others to develop bias reducing strategies. Those can include techniques to help an individual reframe and counteract stereotypes, and consider individuals from their unique perspectives, Kaatz explains. "They remind you that you'll be likely to be making a decision that could be influenced by stereotypes and they help you sort of control that."
Declining the opportunity
Fighting these unconscious biases could also potentially help with another way women are getting lost as peer reviewers. Lerback and Hanson found that women were also more likely to decline invitations to conduct a peer review than men, although this difference seemed to have less of an effect than authors' and editors' recommendations.
Although female scientists turning down the opportunity to be a peer reviewer often cite busyness – and that could be attributed to higher involvement in teaching and outreach, or family commitments – Lerback suggests that perhaps women simply don't see themselves as experts, even when they are.
Workshops aren't only for targeting biases once scientists are involved in writing, reviewing, and editing scientific journal papers. Lerback adds that university clubs could be a good space to develop bias-reducing strategies in scientists before they even reach that point in their careers. And her university is working to launch just such a club that will be geared toward addressing all biases, not just gender biases.
Some journals, like Conservation Biology, Nature Geoscience, and Nature Climate Change, have tried to eliminate potential bias in the peer review process itself by instituting double-blind reviewing systems. In such a scenario, reviewers would be "blind" to the identity of the authors.
Although research has suggested that double-blind reviews can help increase female authorship of scientific papers, it has its downsides, too, as a solution.
One of the main perks of peer review is the networking and recognition it gives both authors and reviewers, Hanson says. As an author and reviewer's respective research is usually very similar, it is a way for scientists to connect.
And, Lerback says, "I think that it's treating the symptoms of a problem rather than the source. The source of the problem is the bias, so if you're just covering that up, you're not addressing where it's actually coming from."
That sort of investigation and transparency around it is the first step, Kaatz says. So the AGU is on the right track.
It's not all uneven for women in the AGU. The study also revealed that female first-authored papers were more readily accepted by an AGU journal (at 61 percent) than those led by male scientists (57 percent). This could be because female scientists work harder on their papers before submitting them for publication, Hanson and Lerback suggest, perhaps as a result of having more challenges throughout their careers.
Still, high publication rates could be a good thing for female scientists, Wendy Williams, director of the Cornell Institute for Women in Science and a professor in the department of human development at Cornell University, writes in an email to the Monitor. Scientists at leading research institutions care little about how often their colleague's peer review papers, "but they care deeply about how many papers they had accepted in journals."