Friday, January 14, 2011

The Stereotype Threat. Or Priming Gender

This is the sixth post in my series about the science of sex differences, and the first one not really about gender differences which some argue to be innate and unalterable sex differences.

Instead, this post discusses the consequences of gender stereotypes, both correct ones (in the sense of averages) and incorrect ones, on the actual performance of girls and women in various tests, in schools and colleges and at work. Studies of sex differences have an impact on sex stereotypes, as we all know. Those sex stereotypes, in turn, can affect the ability of a person to perform as well as she or he can. The way this happens is through something called stereotype threats.

Wikipedia defines the stereotype threat as follows:

Stereotype threat is when a person who belongs to a group that has a negative stereotype attached to it subconsciously conforms to the negative stereotype by performing a task to a lesser degree than they would otherwise.
It turns out that stereotype threats can be created quite rapidly. All it may take is a test-giver's initial announcement that a particular subgroup in general fares poorly/well on that particular test. Stereotype threats exist or can be created about race or ethnicity, and it is quite possible to create a stereotype threat which affects men/boys rather than women/girls, as was shown in my post about three-dimensional mental rotation.

The important word in the above definition is "subconsciously." Cordelia Fine reports on the many ways this priming happens in the context of gender (Delusions of Gender, pp. 7-8):
Some psychologists refer to whatever self is in current use -- the particular self-concept chosen from the multitudes -- as the active self. As the name implies, this is no passive, sloblike entity that idles unchanging day after day, week after week. Rather, the active self is a dynamic chameleon, changing from moment to moment in response to its social environment. Of course, the mind can only make use of what is available -- and for each of us certain portions of the self-concept come more easily to hand than do others. But in all of us, a rather large portion of the Wardrobe of Self is taken up with the stereotypical costumes of the many social identities each person has (New Yorker, father, Hispanic American, vet, squash player, man). Who you are at a particular moment -- which part of your self-concept is active -- turns out to be very sensitive to context. While sometimes your active self will be personal and idiosyncratic, at other times the context will bring one of your social identities hurtling towards the active self for use. With a particular social identity in place, it would not be surprising if self-perception became more stereotypical as a result. In line with this idea, gender seems to have exactly this effect.
In short, if something reminds a woman of her gender while she is undertaking a task in which women are regarded as less capable, her own negative gender stereotypes might be activated.

Why would this matter? Activating stereotype threats may cause physiological stress reactions, reduce working memory capacity or even create a disruptive mental overload. Or to give you an example, when you work into a math exam room and someone yells at you "Hey, token tits!", not only might you have trouble settling down and focusing on your exam paper because of your overt anger, but your subconscious self may also be busy filling up your working memory with stereotype crap while pumping up your blood pressure to cope with the threats in the situation.

Some researchers argue that it's the very activity of trying to repress the negative stereotypes that causes the lower performance of individuals once the stereotype threat has been activated. Some part of the person's brain has to battle the stereotypes, to keep them submerged, and this battle consumes energy which is then not available for thinking about the questions in the test.

Finally, an activated stereotype threat may change the test-taker's attention from a focus on seeking success to a focus of failure-prevention. The latter approach means being cautious, conservative and careful. Astonishingly enough, this behavior would also produce the thinner tails of many female test score distributions, something I discussed in the previous post.

The dampening effect of gender stereotype threats on women's and certain minorities' test performance is now well known from many studies. Stereotype threats exert an independent effect on the performance of the members of the group with negative stereotypes. Though not all individuals are equally susceptible to, say, gender stereotypes, their impact is enough to affect the average performance women and girls in various tests. Cordelia Fine, in Delusions of Gender, discusses many such findings in the first three chapters.

She also points out that stereotype threats are not only elicited by formal testing situations with gender or racial priming. Women who work in male-dominated fields may face stereotype threats on an almost routine basis, especially if they are the lone women in their departments, the ones who have to "represent" the whole female sex in various arguments, the ones whose whole behavior is interpreted as proof of "what women can't do." To the extent such stereotype threats are long-term, they may even explain why some women leave fields such as engineering after a while. It gets tiring to have your blood pressure rise or your working memory decrease because of "disruptive mental loads", as Wikipedia describes the effects of gender priming.

Oddly enough, stereotype threats may be more powerful when they are subtle. Subtle reminders might pass our conscious brain and dive straight into the subconsciousness, whereas we might spot and question coarser stereotypes.

And resisting the stereotype threat doesn't really work, at least if the resistance takes place during the test itself.

This is quite dismal, right? The reason I write about it is that gender stereotypes are created, and the findings from various studies about sex differences certainly contribute to that creation!

To the extent those findings are flawed or biased, to the extent similarity studies are not published and to the extent popularizers let fly with any study which seems to prove the existence of gender differences as innate and unchangeable, it is to that extent that new and possibly false stereotype threats are created.

Thus, bad research in the field of sex differences may have real world consequences. Indeed, if bad research changes the gender stereotypes sufficiently, the new changed stereotype threat could alter reality to match the flawed initial findings!

It is for this reason that any study of sex differences should be carefully scrutinized and even more carefully popularized.

The current practice is the very opposite of that. ANY real or imaginary sex difference is instantaneously plastered over newspapers and web pages, with hyperbolic summaries of the research findings. Indeed, that would be the very way to manufacture stereotype threats if they didn't already exist.
This post is largely based on Cordelia Fine's Delusions of Gender, chapters 2 and 3.