It’s Women’s History Month in the U.S., yet women have nothing to celebrate when it comes to their continuing inequality in the workplace. A recent report from the women’s advocacy group, SPARK Movement, looked at Google doodles over the course of four years, and found that only 17.5 percent of the homepage animations featured women. The report set off a firestorm, as the findings fell in line with the fact that most people who work in the tech haven are men. Now, a new study finds that stereotypes are what kill women’s opportunities for these jobs, which require skills in math and science.

The findings, published in the Proceedings of the National Academy of Sciences, show how an unconscious bias in “hiring managers” can influence them to choose men over women, even when women perform equally well or better on tests, and speak honestly about their skills. The researchers found that these stereotypes, and subsequent hiring practices, held true when the “hiring managers” were both men and women.

Biases Against Women Persist

For the experiment, researchers split the participants into two groups: hiring managers and potential employees. The candidates then completed a four-minute arithmetic test that involved adding up sets of two-digit numbers as fast as possible. This particular test was chosen because both men and women tend to do equally well.

After they completed the test, the hiring managers were tasked with choosing who they wanted to employ. They were given limited information — sometimes only seeing a photo of the person — but were inclined to choose whoever they thought performed best, as there was a higher financial incentive for top-performing candidates. In this exercise, they were twice as likely to hire a man, The New York Times reported.

In another experiment, job candidates were allowed to speak about how they did on the test. While men were more likely to exaggerate, women were more likely to underrate their talent — obviously not a good thing. Because of that, women were once again 50 percent less likely than men to get hired. These biases held up, with only a slight reduction, when the hiring managers were shown hard data about the candidates’ abilities. Because of this, two-thirds of the time that they chose a lower-performing candidate, they chose a man.

The researchers also performed an Implicit Association Test on the hiring managers, which looks at potential unconscious biases by how people associate certain words. Associating “man” rather than “woman” with “math,” for example, points to a possible bias, Science reported. They found that participants who were more prejudiced were less likely to correct their bias in the hiring setting. “The very people who are biased against women about math, they’re also less likely to believe that men boast,” Luigi Zingales, of the Booth School of Business at the University of Chicago, told The New York Times.

Gender Discrimination at Work

The findings have widespread implications, as many women may feel shut out from science, technology, engineering, and math jobs, also known as STEM fields. Whether they’re shut out or face discrimination in the workplace, gender bias affects both morale and motivation. And, according to an August Gallup poll from the 15 percent of women in the U.S. felt that they’d been unfairly passed up for a promotion. During the interview process, many women are also asked about family responsibilities, despite the law prohibiting such questions — they can be turned down if the employer feels that they aren’t capable of managing both home and work responsibilities. According to Mahzarin Banaji, a psychologist at Harvard University, it essentially comes down to “members of disadvantaged groups … costing themselves,” he told Science. That needs to stop.

Source: Reuben E, Sapienza P, Zingales L. How stereotypes impair women’s careers in science. PNAS. 2014.