The decision to pursue a career rests in part on how we judge the following inequality:

Visualization of My Abilities being greater than or equal to the abilities required for success in this career

If we believe this inequality to be true, we might proceed1; if we decide it’s false, we might look elsewhere. Importantly, however, neither side of this inequality is easy to evaluate. Abilities are nebulous, context-sensitive things that are notoriously problematic to pin down. As a result, we often look to others for clues, leaving the door open for substantial social and cultural influences on career choices. A symposium at the 2014 SPSP conference in Austin highlighted a number of recent findings that link sociocultural influences on people’s assessment of the inequality above to the presence of gender gaps.

How do we get from sociocultural influences on this formula all the way to gender gaps? First, and most obviously, contemporary culture is rife with stereotypes about differences in men’s and women’s cognitive profiles; these stereotypes shape people’s beliefs about the quantity on the left-hand side (that is, the abilities they are likely to possess). Second, and less often discussed, practitioners of different careers may send different messages about the abilities that are required to reach the highest levels of achievement in their particular field; these messages shape people’s beliefs about the quantity on the right-hand side (that is, the abilities required for success). Putting these two elements together, we might make the following claim: One circumstance that gives rise to a gender gap in a career or discipline is when a gender group is stereotyped as lacking an ability that the people in that discipline believe is essential for success.2

The four talks in our SPSP symposium examined this idea from a variety of angles. Together, they made a compelling case for people’s beliefs about ability (and the sociocultural influences on these beliefs) as a key factor in the emergence of gender gaps.

In the first talk, Catherine GoodJennifer Mangels, and Laura Deering focused on a domain with pervasive stereotypes against females—mathematics. They asked: Would manipulating students’ beliefs about who belongs in math—about the abilities that are valued in this field—reduce the detrimental effect of stereotypes on middle-school girls’ math engagement and learning? To answer this question, Good and her colleagues randomly assigned 242 7th graders to one of three conditions that differed only in the belonging beliefs conveyed to students. In one condition, children read a brief “news article” that presented “evidence” suggesting that, in order to truly belong in math, one needs to “be a natural.” In another condition, the necessary ingredient for belonging in math was instead said to be hard work. Finally, the remaining children were assigned to a no-information control condition. Across two testing sessions, the researchers assessed students’ math performance, their engagement with math, and their learning from the first to the second test session.

Mind the Gap logo

The second and third talks (featuring research by Sarah-Jane LesliemyselfMeredith Meyer, and Lin Bian) examined the influence of ability beliefs across academia. Because these two talks described aspects of the same research program, I will summarize them together. The starting point for our argument is that the pattern of women’s involvement in academia is more complex than is usually suggested. (Typically, the message is “there are few women in STEM and many women in the social sciences and the humanities.”) For example, gender gaps are far more dramatic in some STEM disciplines (e.g., physics) than in others (e.g., biochemistry). In addition, there are disciplines in the social sciences and humanities with gender gaps as large as some of those in STEM (e.g., economics and philosophy). We argued that attending to these complexities—specifically, to the variability in women’s representation within STEM, as well as the sizable gender gaps in some non-STEM disciplines—may provide new insights into the causes of women’s underrepresentation across academia. We then proposed a hypothesis that might account for this complex pattern of underrepresentation. Specifically, we proposed there is a powerful feature shared by all disciplines, regardless of whether or not they are in STEM, that predicts their female representation: namely, their practitioners’ beliefs about what is required for success. Just as individuals differ in their personal opinions on this topic (as work by Carol Dweck and her colleagues have compellingly shown), so might entire disciplines: In some disciplines success may be seen as depending on sustained effort and dedication, whereas in others it may be seen as requiring a “gift” or brilliance that cannot be taught. Because women are stereotyped as being less likely than men to possess innate intellectual talent, they may find the academic fields that emphasize brilliance as the key to success to be unwelcoming. (In case anyone doubts that the stereotypes against women’s intellectual abilities are still with us in the 21st century, they should take a look at a New York Times article from a few months ago in which the author analyzed Google searches and found that parents were 2.5 times more likely to ask Google if their sons were gifted or geniuses than if their daughters were.)

To test our hypothesis, we first conducted a nationwide study of about 1,800 academics across 30 disciplines and found that—as predicted—the more a field emphasized brilliance, the lower women’s representation at the PhD level. (Stunningly, the correlation was as strong as −.60.) Moreover, this relationship held when looking at STEM and non-STEM fields separately, as well as when a number of other competing variables were taken into account (e.g., the average number of hours worked by people in a field). We followed up this finding with a series of experiments that made the same point: When we introduced a new opportunity to participants (e.g., an internship, a major) and described it as requiring brilliance, talent, etc., women were less likely to want to pursue it relative to men (whereas this difference was absent or reversed when we described the same opportunity as requiring dedication, effort, etc.). Much to our dismay, we found exactly the same pattern in 6- and 7-year-olds. Girls were less likely than boys to want to play a new game that was said to be “for kids who are really smart,” but there was no difference between boys and girls when the game was said to be “for kids who try really hard.” Again, this research clearly illustrates the role of sociocultural messages about (1) one’s likely abilities (here, in the form of gender stereotypes) and (2) the abilities required for success (here, in the form of discipline-specific messages about requisite skills). When combined, these two elements seem to provide a powerful tool for understanding gender gaps.  

Finally, the talk by Joseph P. Robinson-CimpianSarah T. LubienskiColleen M. Ganley, and Yasemin Copur-Gencturk provided a different sort of evidence for the idea that societal beliefs about ability may give rise to gender gaps. Using a nationally representative dataset collected by the Department of Education, their research suggested that (1) teachers hold stereotypes against girls’ math ability, and (2) these stereotypes account for a substantial portion of the development of the math gender gap in the early elementary school grades. What’s the evidence that teachers hold negative stereotypes about girls? When you look just at the raw means, teachers’ ratings of girls’ math competence are actually higher than those for boys—so there seems to be no evidence of stereotyping there. However the comparison of raw means is misleading because boys and girls differ on a number of other important dimensions that need to be taken into account (e.g., teachers’ impressions of girls’ and boys’ persistence and engagement). If you adjust for these variables, you get the opposite picture. That is, when you look at boys and girls who are matched on their persistence, attentiveness, etc.—as well as their actual math performance—teachers in fact rate the boys as being more competent in math. (I should also mention that the measure of math performance used to match boys and girls was custom-made for the Dept. of Education and was quite detailed and comprehensive.)

Next, Robinson-Cimpian and his colleagues examined whether this underrating of girls influenced the development of the math gender gap. With some complex statistical techniques that allowed them to approximate causal conclusions with observational/correlational data (instrumental variables and propensity score matching), they found that teachers’ underrating of girls may have been responsible for as much as 80% of the growth of the gender gaps from kindergarten to 1st grade to 3rd grade. (The estimates vary depending on the statistical model you look at—I selected the highest to illustrate the potential power of these stereotypes in shaping the development of gender gaps. None of the estimates are lower than 45%.)

At this point, some readers may be thinking: Isn’t there a fact of the matter here? Don’t men and women actually have different cognitive skills due to biological differences? Don’t various fields actually require different amounts of intellectual talent (or different talents)? Questions about actual differences in abilities are complex and mired in vitriolic debate. (But no—I have seen no convincing evidence of biological differences in men's and women’s cognition that would be relevant to success in intellectual pursuits of the sort we care about.) Importantly, however, the mechanism I described here is likely to operate no matter what we find out about these other issues. Regardless of the purported cognitive differences between men and women, or of the abilities purportedly required to become a physicist vs. a psychologist vs. an anthropologist, the mere presence of (1) different societal beliefs about the intellectual abilities of men and women, and (2) different societal beliefs about the intellectual abilities required for success in different fields will be sufficient to give rise to (or at least exacerbate) gender gaps. At its core, this is an optimistic conclusion: Intervening to change the relevant beliefs about ability (and how they are projected to children and students) is likely to bring about greater gender equity. It won’t be easy to make such changes, of course, but the end goal is certainly worth it.


Andrei Cimpian is a developmental psychologist at the University of Illinois at Urbana-Champaign. His research focuses on the development of human concepts about natural and social categories, as well as on issues pertaining to academic achievement (e.g., stereotypes, and people’s beliefs about ability and talent). You can find more information about his work on his lab’s website and on Twitter, @AndreiCimpian.

  1.  Assuming, of course, that we are also interested, motivated, and so on.
  2. This claim applies to other groups as well (e.g., groups based on race/ethnicity) rather than being specific to gender.