Peer-reviewers twice as likely to accept research conducted on men than the same research on women
How biases may play out against female experimental participants.
11 May 2021
Research with male experimental participants is almost twice as likely to be accepted during peer review as the same research conducted with women, a new study finds.
The authors of the study, published in the Journal of Women’s Health in March, assigned peer-reviewers one of three versions of study abstracts: either a version saying the research was conducted on men, one saying the work was conducted on women, or a third type simply referring to “individuals”, rather than either gender.
As a cover story, the authors told the study participants that they were participating in research for a new journal that was investigating whether robust peer review can take place by evaluating only abstracts instead of whole manuscripts.
Out of the 17,296 potential peer-reviewers the authors contacted — all of whom were recipients of grant funding from the US National Institutes of Health and were listed in the agency’s RePORTER database — only 315 supplied responses. The study controlled for respondents’ characteristics, such as age, gender and race.
The participants were found to be significantly more likely to consider research with female subjects as a greater contribution to medical science than the same research on men. Even so, they were twice as likely to recommend studies on men for publication.
When asked which studies were more scientifically rigorous, the survey participants rated the papers of similar rigour across genders. It's remarkable then, that they still favoured work conducted with male participants when it came to publication, says study co-author Sohad Murrar, a social psychologist at the Governors State University in University Park, Illinois.
“Whether it is conscious or unconscious, this bias results from the pervasive knowledge of gender stereotypes and the social norms that are constantly being reinforced by the existence of these stereotypes,” says study co-author Molly Carnes, a physician and geriatrics researcher at the University of Wisconsin-Madison.
Bias against studies on women resulting in lack of data about them could have “huge ramifications” for women’s health and wellbeing, says Murrar. “I think that we might be missing very important information that needs to get out there about women’s health as a result of these types of biases that might exist.”
Calls for further investigation
Cassidy Sugimoto, an information scientist at Indiana University Bloomington, who has written extensively about gender disparities in science, says she was excited to see that the new study was a randomized control trial (RCT), which is widely considered to be the gold-standard technique in figuring out whether a trend truly exists.
A key tenet of RCTs is a control variable; in the new study, this was the category of studies which mentioned subjects as “individuals” rather than specifying their gender.
For Sugimoto, however, the new analysis doesn’t conclusively prove that there are biases against studies with female participants. This is because the sample papers Murrar and colleagues used in their analysis were about sexual trauma — a research topic that’s heavily skewed towards women as subjects.
“It’s problematic when you have any topic that has either perceived or known skew towards a certain gender,” Sugimoto says. She adds that it’s unclear whether reviewers were more keen to accept papers related to men because sexual trauma in men is not studied as frequently.
Mathias Wullum Nielsen, a sociologist at the University of Copenhagen in Denmark, echoes Sugimoto’s concerns. Both Nielsen and Sugimoto emphasized that follow-up studies with larger sample sizes are crucial to figuring out whether biases really exist against female experimental participants.
“When you exclude studies on women,” Sugimoto says, “that also implies that you’re excluding studies by women, because women are more likely to be doing studies with female subjects.”