The Why Files The Why Files --

Women in science: What are the obstacles?

Science: A social matter?
If women were genetically unable to do science, how have they advanced so quickly in the past 30 years?What is the truth about women in science? Are women a bit slower than men at the top end, or do discrimination and bias explain why men predominate at the highest reaches of academic science?

You may recognize this as a reprise of the long debate over the roles of genetics and upbringing: Are you smart because you have good genes, or because you had those stimulating early-childhood toys and a great education?

This "nature versus nurture" is still raging after more than a century:

Do we speak because our brains are wired for speech, or because language is useful and we hear people talking from day one?

Do some students excel in math and science because they are smart, or because they work hard?

Do you get tenure at Harvard because of innate brilliance, or because of a lifetime of social and academic advantages?

Now rewind to Summers's original explanation for why women don't get as many key jobs in science. Number one was their supposed reluctance to work "80-hour weeks." Number two was "The different availability of aptitude at the high end." In other words, brilliant women aren't quite as science-smart as brilliant men.

Lastly, Summers named socialization -- upbringing -- as playing the least role in hindering women in science. Genetics explains why men do better than women; nature trumps nurture.

No way, say many women scientists we talked to. For one thing, they ask how women could have advanced so fast in recent years, even in physics, a quintessentially male field. That doesn't sound like genetic change, says Caitilyn Allen, who teaches plant pathology and women's studies at the University of Wisconsin-Madison. "When you are trying to drag out a biological explanation, if women were profoundly disabled at a genetic level, it would be hard to reconcile with the facts. In 1970, 0.3 percent of undergraduate engineering degrees went to women, and now it's 17 percent. It would be hard to argue that women have improved their quantitative skills 51 times in 35 years. Genetics does not work that way, and that makes it clear that there is significant societal pressure that affects women's success in science."African-American woman scientist grins as she peers at a Petri dish.

Rather than overt discrimination, Allen blames "accruing small differences over time. It's rarely a matter of overt, blatant discrimination. That used to be the case, but it no longer is."

Courtney Robinson, microbiology PhD student, University of Wisconsin-Madison: "I think [discrimination] is definitely something you face as a woman. As an African-American, I've faced some of the same questions, but they come from a different source. I'm in a position here at Wisconsin, where I don't feel I have to prove myself as a woman, but I do feel that I have to prove myself as an African-American. Here I'm an African-American before I'm a woman."

What does this mean?
Discrimination does seem to emerge from experimental studies that look at how women and men judge men and women. For example, says Allen, "one study found that if you sent a paper out for review and identified it as being written by John Doe or Joan Doe, you get a different response... The paper with the female author was significantly less well regarded by the reviewers."

That test is called the "Goldberg paradigm," and variations have been done repeatedly: Randomly assign a man's or a woman's name to some articles, and ask people to judge the articles -- and their authors. Several women scientists we spoke to mentioned that the studies prove that women's work is judged more harshly, simply because a female name appears as author.

But a 1989 analysis of this literature (see "Joan McKay versus John McKay ..." in the bibliography) found something slightly different, and perhaps more interesting. In that article, Janet Swim, who's now resides in the Penn State psychology department, wrote, "In summary, little evidence was found for the simple prediction that subjects differentially evaluate men and women in the context of the classic paradigm that Goldberg introduced in 1968."

Graph shows increase in number of physics degrees awarded to women.Swim told The Why Files that her article is credited for finding bias, when in fact it found only a tiny amount. "If you rely entirely on the Goldberg paradigm, without any thought about when it will and when it won't apply, it's too simple."

Over the past 30 years, women have made great strides in physics degrees. Graph: AIP

Chieh-Chen Bowen, an associate professor of psychology at Cleveland State University, explains that meta-analyses combine research from "different settings, different subjects, they might use different targets for rating, and all those have an effect on the result. A lot of times, when you integrate large numbers of studies, positive and negative studies tend to cancel each other out."

So even though people do meta-analyses because they expect to find a relationship between two phenomenon, "You have to report what the data says," Bowen says. "If you read Janet Swim ... you have to conclude gender would not affect the evaluation of men's versus women's work."

Whatever happened to the real life?
Bowen has pursued the question of how men and women evaluate the work of women and men, but has changed the focus to "real job performance across different settings" rather than the laboratory studies that Swim analyzed. For that reason, her results may reflect job performance, bias, or both. "Maybe their actual performance is different, but I could not tease that out" (see "Evaluating Gender Biases ..." in the bibliography).

In terms of bias, she told us, setting matters. "In law enforcement ... they don't like women; women's work is valued much lower," she says, compared to fields like education. While the data vary by the setting, the overall results are similar to Swim's, Bowen says. "If you look at everybody together, they average out, there is no effect."

Still, individual studies do reveal circumstances where the work of women is likely to be devalued, she adds. "If the person is talking about finance, there is probably not much difference, but if it's about science or engineering, all of a sudden the male author gets a higher rating for clarity, accuracy, for contribution to the field."

What else do we know about subtle bias?


Megan Anderson, project assistant; Terry Devitt, editor; S.V. Medaris, designer/illustrator; David Tenenbaum, feature writer; Amy Toburen, content development executive

©2023, University of Wisconsin, Board of Regents.