Eye of the Beholder

When it comes to judging evidence of gender bias in the sciences, is gender bias at play? A study involving university faculty members suggests it is.

October 13, 2015

More and more research suggests gender bias in the sciences. But do men and women similarly trust evidence demonstrating such bias? A new paper argues that men and women interpret this kind of evidence -- however scientific -- differently, and that that has implications for the field as a whole. Perhaps unsurprisingly, though, not everyone agrees with the findings.

The paper, in the current Proceedings of the National Academy of Sciences, is based on several experiments involving both laymen and -women and male and female faculty members. All participants were asked to evaluate scientific research showing a bias against women in science, technology, engineering and math. Men in both groups judged the research less favorably than did women, and male faculty members in STEM were especially likely to view it poorly. But men evaluated the same research more favorably than did women when findings of that research were altered to show no gender bias.

“Critically, across three experiments, we uncovered a gender difference in the way people from the general public and STEM faculty evaluate the quality of research that demonstrates women’s documented disadvantage in STEM fields: men think the research is of lower quality, whereas women think the research is of higher quality,” the authors said. “Why does this gender difference matter? For one, there are significant implications for the dissemination and impact of meritorious previous, current and future research on gender bias in STEM fields.”

Most importantly, they argue, “our research suggests that men will relatively disfavor -- and women will relatively favor -- research demonstrating this bias.” That means that scholars focusing on this area of research might have trouble advancing professionally, given that men still dominate the STEM fields, they say, and that lingering doubts about bias in STEM will unnecessarily prolong the process of making the sciences more inclusive.

Ian M. Handley, a professor of psychology at Montana State University, served as lead author of “Quality of evidence revealing subtle gender biases in science is in the eye of beholder.” The study involved two general U.S. population samples of a 205 and 303 people, respectively, and a Montana State faculty sample of 205. Using an online survey instrument, Handley and his collaborators presented a general U.S. sample and the faculty sample an abstract of a paper that reported gender bias in STEM contexts. They then asked participants to evaluate the strength of the abstract and findings.

In a third experiment involving a second U.S. general sample, the researchers asked participants to evaluate the abstract of a different article suggesting gender bias in STEM fields. Members of that group received either the real abstract or a version slightly altered to report no gender bias.

In the first two experiments, the difference between men’s and women’s evaluations was of a moderate size. But the difference was larger between male and female STEM professors. Interestingly, no significant difference was observed between male and female non-STEM faculty members. Put another way, male STEM professors evaluated the research more negatively than did the non-STEM male professors in the sample.

In the third experiment involving a second abstract, men were somewhat more likely than women to evaluate positively the altered abstracts showing no bias.

“Much of the research on the issue demonstrates a bias against women in STEM, and our research suggests that men, including faculty, are less receptive to that evidence than women,” Handley said.

Yet the bias can go both ways, he added. “Our research also demonstrates than women are less receptive than men to research that demonstrates no gender bias in STEM contexts.”

As academics, Handley said, it’s “useful to understand that we are prone to bias regarding some research findings, and our objectivity going forward will require this acknowledgment.”

Handley’s co-authors are Elizabeth Brown, an assistant professor of psychology at the University of North Florida; Corinne Moss-Racusin, an assistant professor of psychology at Skidmore College; and Jessi L. Smith, a fellow professor of psychology at Montana State.

“Eye of the beholder” is similar in theme to an earlier paper by Moss-Racusin suggesting men were much more likely to reject findings of sexism in STEM and to even make sexist comments in response to such research. That study was based on Internet comments on various kinds of websites regarding a third, widely cited study by Moss-Racusin suggesting that both male and female scientists were more likely to want to hire as a lab manager and award higher pay to a hypothetical student candidate named John than one named Jennifer, even though the rest of their applications were identical.

“Eye of the beholder” is arguably a more informative paper on gendered evaluations of gender bias research in academe than the one based on Internet comments, in that it involves double-blind experiments and university faculty specifically. Still, it has some critics. Among them are Wendy Williams, professor of human development at Cornell University, and Stephen Ceci, the Helen L. Carr Professor of Developmental Psychology at Cornell, who together have authored several papers questioning the existence of gender bias in STEM. Most recently, earlier this year, they published a widely cited paper suggesting that women candidates are actually favored two to one over men for tenure-track positions in STEM.

Williams said that, due in part to her findings, men evaluating evidence of gender bias in STEM might have reason to be skeptical -- meaning that their negative perceptions might not be a form of discrimination but rather an awareness of the literature.

Ceci agreed. But he also noted it’s possible that the Moss-Racusin lab manager hiring study -- which served as one of the abstracts in “Eye of the beholder” -- may have demonstrated true bias when it comes to the hiring of recent graduates. That's because faculty members may rely on stereotypes when hiring candidates whose professional aptitude is relatively untested, he said. Ceci’s own research showing no or even reverse gender bias in hiring, on the other hand, focuses on “unambiguously qualified” Ph.D.s seeking tenure-track jobs.

Both Williams and Ceci strongly cautioned against applying Handley’s observed results, involving a 205-member faculty sample from just one university in one part of the country, to academe as a whole.

“Our study looked at 873 faculty from 371 universities and colleges in all 50 U.S. states,” Williams said. “Would these results have replicated in a broad national sample that included a few hundred colleges and universities” in different parts of the country, she asked?

Handley rejected the idea that a 205 faculty member sample was too small to observe a significant trend. Moreover, he said, the general population sample was larger, involved people from across the U.S. and showed similar results.

“Faculty are people too, and our experiments on the general population suggest the effects are likely generalizable,” he said.

Smith, the study's senior author, said she believed the results would replicate at other universities, but that she hoped they wouldn't replicate at Montana State today, which has worked hard to make the sciences more inclusive in the last few years, since the faculty survey was conducted. (The university recently received an award from the College and University Professionals Association for Human Resources for its efforts.) In any case, Smith asked, "What is worse – understating a bias that exists which results in keeping people from fully participating in STEM or overstating a bias that exists which results in real transformation and resources to correct a past injustice?"

She added, "There are so many calls for research on broadening participation in STEM and for research providing evidence-based solutions to bias, that it is important to understand how receptive people are to such findings. How can we solve a problem if we don’t know we have it?"

Handley said he was most interested why these results occur in any sample. One reason might be confirmation bias, he said, in which people are more likely to evaluate favorably information that corresponds to their own beliefs or experiences.

Of course, there may be some campuses on which men and women largely agree on the pervasiveness of gender bias in STEM, Handley said. “I certainly acknowledge that there is more to this story, which means there is more research we can do to assess the pervasiveness of this phenomenon, and how to reduce it in the efforts of enhancing objectivity over all.”


Back to Top