In Making Hiring Decisions, Men and Women Still Assume Women Are Bad At Math

In an experiment on bias, women who aced math problems were more often overlooked in favor of male applicants who performed worse.

In Making Hiring Decisions, Men and Women Still Assume Women Are Bad At Math
[Image: Businesswoman via Shutterstock]

One of the most stubborn and pervasive gender-based stereotypes out there is that women are bad at math. I’ve heard versions of this pseudo-scientific crap voiced by technical and non-technical people alike, by stone-cold sexists and friends I respect, too. It’s odd then, that for people who supposedly just “state the facts,” even hard numbers disproving those facts are apparently unlikely to dislodge their biased assumptions.


That’s at least the finding of a study recently published in the Proceedings of the Academy of National Sciences. The research examined the assumptions that hiring managers make when considering male and female employees for a math-related task. But while much of what the researchers found was discouraging, the authors also suggest that one of the tests they used to assess an employer’s bias could be deployed to fix that error in real-life hiring decisions.

So, the bad news: When test subjects assigned to be the “employer” had no information about their prospective employees’ abilities, they were twice as likely to choose men over women. In another scenario, where both male and female candidates were given the opportunity to describe how well they thought they would do on the math task, the outcome wasn’t much different.

Things get more interesting in the third scenario, in which employers could actually “update” their decisions after watching men and women do the math problem. Employers still only picked women 43% of the time. More tellingly, the “employers” picked the job candidate who performed worse 19.5% of the time. Of the worse performers chosen over better mathletes of the opposite gender, 64% were male.

This outcome, researchers say, has much to do with gender bias. “If hiring decisions were sex-neutral, the fraction of sub-optimal decisions in which a lower-performing male was chosen over a higher-performing female would be close to 50%,” the researchers write. Instead, men who performed worse on the math test more often got hired anyway.

But what if employers were made aware of their bias?

The researchers noted that sexist hiring decisions corresponded with scores on something called an “Implicit Assumption Test,” a tool originally developed to test racial bias. The IAT, tweaked to assess gender bias, works like rapid-fire flash cards: The test subject is shown a word or image in the center of the screen, and is asked to associate that object with the right side “humanities” or left, “math and science.” Your prejudice is then calculated by your response–if you lump “Sheila,” “Katie,” and “Veronica” into humanities while consistently pushing “Michael,” “Roy,” and “Dave” into math, you’d receive a positive score for associating men more with math and science, and women less. If it worked the opposite way, you’d receive a negative number.


The paper found that both men and women employers had positive mean IAT scores: 0.35 for men and 0.42 for women. Unsurprisingly, employers with higher IAT scores also rated men’s self-assessments more highly, even though, researchers discovered, they were often far more overblown (inaccurate) than women’s.

Ernesto Reuben, one of the authors of the study, suggests that hiring managers could actually use the IAT to assess their own prejudice when making decisions.

“We do see that discrimination does go down [when employers are given the facts about performance], but they just don’t go far enough,” says Reuben. “They put too much weight on this stereotype, which actually has very little informational content.”

Still, Reuben added that it might be unethical for any company to mandate something like this–and if you want to avoid sexist labeling, you also might be able to game the test by simply being slow and “associating” neutral answers. At the same time, it could be a critical personal diagnostic to use, as Reuben suspects that lack of women in science, technology, and engineering fields could be the very thing contributing to the power of the stereotype. Plus, it might mean that people with high IAT scores consistently hire less qualified people.

“My personal take is that a lot of this is just unconscious processing by the brain,” Reuben says. “If you never see a woman in math-related fields, your brain automatically links men with math, and women with other things, and that’s what perpetuates this effect.”

About the author

Sydney Brownstone is a Seattle-based former staff writer at Co.Exist. She lives in a Brooklyn apartment with windows that don’t quite open, and covers environment, health, and data.