Stereotype Threat—What STEM can learn from Social Sciences

By Tooba Fatima

I was invited recently to speak at a seminar on making STEM (Science, Technology, Engineering, & Mathematics) education more accessible for women, arranged at the UET, Lahore. As the best ideas come from sharing, I thought I’d bounce ideas off friends and colleagues. One of the immediate responses I heard was, it’s really not that big of a problem anymore. I mean, just from Pakistan a woman has won the Oscar (twice), two women have been affiliated with the Nobel Prize for peace and physics, and the women’s team has objectively outperformed the men’s team this year.

This line of argument captures two of the worst mistakes you can make:

  1. Using exceptions to define the norm; and
  2. Pretending that a problem doesn’t exist based on the prominent, but, rare exceptions.

The reason we need to continue conversations about gender is not because men are evil. But because unlike men, women are systemically oppressed. There are systems—economic, psychological, cultural structures bigger and more powerful than any individual man or woman—that systemically deprive women of their rights and diminish their status.

When we do turn our eye from the exception to the norm, the case is grace indeed. Data presented by Alif Ailaan reports that as many as 55% of girls (13.7 million of them) are out of school, and millions more never finish primary school, let alone get to participate in STEM subjects. Surveys of developing countries show that only 30% of total degrees awarded in STEM fields go to women.

Most women are familiar with the feeling of not being enough. Of knowing deep down that the same degree and experience and qualification just won’t cut it. They will always have to push harder, sacrifice additionally, be more before they are deemed equal; a strange oxymoron really.

While the research work I personally study or undertake falls primarily under psychology, all social sciences do pay homage to the scientific method perfected by STEM fields. We follow it as closely as possible to study human and social behaviour. And a wealth of interesting information is being generated that helps us understand what holds women back in STEM fields.

Silvia Galdi and her colleagues (2014) conducted an experiment with 240 6-year old children. Each child was randomly assigned to colour one of three possible pictures—a gender stereotypical picture (such as a girl baking a cake), a gender a-stereotypical one (such as a female mechanic), or a neutral landscape. After the child had coloured their assigned picture, they were asked a series of 8 mathematical questions. Boys performed uniformly across all three conditions. Girls however, performed significantly better when they coloured a gender a-stereotypical picture before their math test, and performed significantly worse when the gender stereotypical one.

In another study by Davies et al. (2002), women who watched TV commercials portraying women in gender stereotypical roles (such as kitchen soap or laundry detergent ads) performed significantly worse on a subsequent math test, compared to women who were shown commercials where a woman might be addressing a social or political issue in an intelligent manner.

In fact, a large body of research indicates that even subtly activating women’s gender identity can worsen their performance in mathematics. Women who have to state (only state) their gender before a math test do worse than women who have to state it after the test is complete. Women given the same math test but told that it is a test of their test-taking strategies perform better than women who are told that it is a diagnostic test of mathematical ability. Asian women perform better in math when the test is preceded by questions about race, than questions about gender; because while being Asian is associated with being good at math, being a woman is associated with being bad at it.

These findings have been replicated not only in the lab, but in natural environments. Right at this time several bright, young students are surely preparing to give the NAT, SAT, or GRE in hopes of qualifying for graduate school. Standardised tests usually comprise of verbal and math sections, where the order of appearance is random for each test-taker. Smeding et al. (2013) analysed data from standardised tests, and found them women who were allotted the math section first underperformed not only in math, but in the verbal section. That is, the threat to one’s sense of efficacy had damaging effects above and beyond math-related performance.

Psychologists have been studying the phenomenon of stereotype threat—an anxiety that one’s performance will confirm a negative stereotype about one’s in group, either to oneself or others. The bias that women are less capable—at math, and in general—is prevalent across the world. We see its reflections each day in our advertisements, popular entertainment, and day-to-day conversations. The most prominent consequence of the anxiety generated by stereotype threat is that it increases activity in brain areas that process social cues, and drains resources from brain areas that process working memory and problem solving. The latter two skills being, of course, instrumental in successfully solving math and science problems.

The research that has identified this problem has also identified solutions. Exposing women to powerful female role models, making them recognise similarities between men and women, and educating them about stereotype threat are all effective ways to improve their performance. Just like the problem though, I believe the real solutions lie in broader structural changes.

There is a poem by Martin Nemollier that goes:

First they came for the socialists, and I did not speak out—

Because I am not a socialist.

Then they came for the trade unionists, and I did not speak out—

Because I am not a trade unionist.

Then they came for the Jews, but I did not speak out—

Because I am not a Jew.

Then they came for me—and there was no one left to speak for me.

 Good societies are made up of people who care about each other. We have to understand that every problem in society is a collective one. Effective solutions will arise not when everyone fights for their own cause, but when we take ownership of each other’s concerns and fight for them as we would for our own.

The second has to do with how we structure education—higher education in particular. Involving more women in STEM fields may be a concern of the sciences, but the reasons for exclusion and strategies for inclusion are something humanities and social sciences are studying extensively. We have to begin to conceive of education as something integrated—following models of liberal arts educations where we can learn from disciplines and hear voices other than our own. So that we produce students who are not only more technically skilled, but better aware of the integrated forces that influence their lives. Professions may be about what we do, but education is about who we are.

The conversation about gender is already taking place—through legislation, media, and public forums. It falls on all of us to take it forward.


The author is a Fulbright Scholar, with a Bachelor’s degree in Psychology and English from Forman Christian College, and works as Senior Manager for Research at Rabtt.



  • Galdi, S., Cadinu, M. & Tomasetto, C. (2014). The roots of stereotype threat: When automatic associations disrupt girls’ math performance. Child Development, 85(1), 250-263.
  • Shapiro, J. & Williams, A. (2012). The role of stereotype threats in undermining girls’ and women’s performance and interest in STEM fields. Sex Roles, 66, 175-183.
  • Smeding, A., Dumas, F., Loose, F., & Regner, I. (2013). Order of administration of math and verbal tests: An ecological intervention to reduce stereotype threat on girls’ math performance. Journal of Educational Psychology, 105(3), 850-860.
  • Tomasetto, C., Alparone, F., & Cadinu, M. (2011). Girls’ math performance under stereotype threat: The moderating role of mother’s gender stereotypes. Developmental Psychology, 47(4), 943-949.
Posted in: