Take a moment and think of a person in a recent movie or show who is brilliant—a person with a keen intellect, capable of leaps of insight. Who comes to mind first?

Odds are that you thought of a man—someone like Dr. House, Sherlock Holmes, or Walter White from Breaking Bad. In popular culture and in people’s minds, brilliance and genius are qualities associated with men more than women. In fact, our latest studies show that this association takes hold surprisingly early, by the time children are 6 or 7 years of age.

Like most stereotypes, the associative link between brilliance and males is psychologically powerful. It acts like a lens that distorts how young people perceive themselves and others, with important consequences for their aspirations for the future. If a group is stereotyped as lacking a certain trait, careers thought to require that trait will seem less attainable to them.

It is thus possible that stereotyped notions of brilliance interfere with women’s participation in careers where success is thought to depend on being brilliant. These careers are also some of the most prestigious in our society, so there is a lot at stake in figuring out if women are being excluded by old-fashioned prejudices.

A study published in the journal Science in 2015 explored this question. We surveyed academics in 30 fields across the humanities, social sciences, and natural sciences. We asked these academics to what extent they thought that success in their field depended on some raw, unschooled talent—in other words, on being brilliant.

We found that fields whose members placed more value on brilliance (fields like philosophy and physics) also had fewer women obtaining Ph.Ds. In fact, a one-point increase in a field’s emphasis on brilliance, measured on a seven-point scale, was accompanied by a 28.6 percentage point drop in female Ph.D.s.

Might it be that some fields simply require more brilliance than others? Our analyses took into account each field’s selectivity in Ph.D. admissions and its applicants’ scores on standardized examinations, including math GRE scores. Over and above these objective markers of aptitude, academics’ beliefs about brilliance predicted lower participation by women. Since brilliance is a relatively nebulous concept in people’s minds, and judgments of brilliance are often complex and subjective, this opens the door even wider to stereotyping and bias against people whom society does not typically consider brilliant.

A later study found the same relationship using a subtler measure of a field’s emphasis on brilliance: the frequency of the words “brilliant” and “genius” on RateMyProfessors.com, where there are millions of anonymous reviews of college instructors. Strikingly, fields in which the words “brilliant” and “genius” were more common on RateMyProfessors had fewer women earning Ph.D.s, and the same was true for bachelor’s degrees.

Why are women underrepresented in fields that cherish brilliance? Our most recent research points to the messages our culture sends to young people about the intellectual abilities of men and women.

In a recent paper in the journal Science, with psychologist Lin Bian as lead author, we reported that the stereotype equating brilliance with males is familiar even to young kids. We asked five-, six-, and seven-year-olds a series of questions to determine whether they associate brilliance (or, rather, “being really, really smart”) with one gender more than the other. For example, we told them a short story about a person who was really smart, leaving out any clues to the person’s gender, and then asked children to guess which of four unfamiliar adults (two men, two women) the story had been about.

We didn’t see any gender differences among the five-year-olds. At this age, boys and girls consistently picked members of their own gender as really smart. However, gender differences emerged among the six- and seven-year-olds in our study, with girls selecting members of their gender less often than boys did. This is despite the fact that girls were also certain that girls do better in school than boys do, once again underscoring that these stereotypes can float free of objective markers of achievement.

We then showed children unfamiliar activities that we said were intended for “children who are really, really smart”—the child-friendly equivalent of careers where brilliance is valued. Although boys and girls were equally interested in these activities at age 5, girls became less interested in them (relative to boys) at the age of six and seven. These differences parallel the changes in children’s stereotypes about brilliance. In fact, girls’ interest in the “smart” activities was lower if they endorsed these stereotypes.

We don’t yet know exactly where kids get these ideas from (parents? teachers? TV shows?). Pinpointing the sources of this stereotype is next on our agenda, but the answer is unlikely to be a simple one. Whatever the answer turns out to be, though, it won’t change the fact that girls’ aspirations may be held back by misguided ideas about the intellectual abilities of men and women. If we work together to dismantle these ideas, as well as the brilliance fetish that pervades many fields, we will do better by our little girls—and by the women they will grow up to be.