Gender, science & stereotypes

Paradoxically, this post on gender stereotypes actually reinforced one perhaps unjust prejudiced stereotype I have - that female/feminist writers are apt to rely on unscientific anecdotes and focus upon trivial everyday irritations. This is unfortunate, because rigorous scientific thinking about stereotypes suggests that women have a genuine grievance.

This paper shows what I mean. Researchers got subjects to compete in a task in which it is thought that men do better - mentally rotating different shapes.

They found that women did indeed do worse than men in such tasks, but only when they knew their competitors' gender:

Information on rival’s gender affects women and men very differently. In the task that is perceived to favor men, it has a positive effect on men’s performance under competition, increasing their performance by almost 60%, but a negative effect on women’s performance when competing, reducing their performance in about 40%.

What's going on here is the stereotype threat, or the "give a dog a bad name" effect. When people are invited to believe that something is "man's work", men step up their effort whilst women become weedy girlies.

Stereotypes, then, are not just irritating generalizations. They have serious and large effects.

I find it plausible that this could have adverse effects upon women's life chances. If some jobs are perceived as "men's work" then women will be less inclined to do them, or might do them worse, even if there is no original rational basis for that perception; it's the perception that creates the reality, not vice versa.

What should be done about this? The answer might not be positive discrimination, such as having quotas for women directors. The authors say:

Affirmative action policies based on gender may in fact have counterproductive effects, since while creating advantageous conditions for women they also make gender information salient, affecting women’s performance negatively.

Instead, I suspect the solution is to challenge the original stereotypes - to show that gender differences either do not exist or, where they do, that they are the result of a social construct, rather than innate biological differences; the work of Alison Booth is relevant here.

Now, it should be obvious from that I say this not to criticize feminists. Quite the opposite. Underneath the trivial anecdotes lies a real and important issue.

Comments

[Tim Newman]

When people are invited to believe that something is "man's work", men step up their effort whilst women become weedy girlies.

Does the same effect occur when men are asked to do what is perceived to be women's work? If so, then this effects both sexes, not just women.

[aliquid]

This is quite interesting, is their any other similar findings or is it just this one?

In other news... COMMMEEEE ONNNNNN WEEESSSTTT HAAAAAMMMMMMM!!!!! IRRONNNNSSSS!!!!

Ahem...

[chris]

@ Tim - the paper reports zero effects for gender-neutral work. I wouldn't be surprised if there is a simlar effect for "women's work" - eg when married men are worse at housework than single ones. But the point is that economically and politicaly powerful jobs tend to be "men's work", with the result that stereotype threat empowers men and disempowers women.

[David Jones]

Does it strike you as at all plausible that there are no between-group differences?

[Andrew]

" On the other hand, in the task that
is perceived to favor women the effect is positive for both men and women"

So, isn't this experiment saying that men compete better than women? That they are not vulnerable to stereotype threat?

Why wouldn't you preferentially employ that trait?

[Torquil Macneil]

One fairly simple way to partially address this problem is a return to education segregated by sex. There is plenty of evidence that women who go to girls' schools have less sexually stereotypical careers.

[redpesto]

"one perhaps unjust prejudiced stereotype I have - that female/feminist writers are apt to rely on unscientific anecdotes and focus upon trivial everyday irritations."

...because the feminists who 'get science' aren't the ones writing about 'unscientific anecdotes and [...] trivial everyday irritations'? (See Ben Goldacre's criticisms of the arts background of a lot of journalists?)

[redpesto]

"There is plenty of evidence that women who go to girls' schools have less sexually stereotypical careers."

Except that doesn't seem to work for boys-only schools, and it's mathematically impossible to have both single-sex schools for girls and a leavening of girls to ensure the boys do well.

[Torquil Macneil]

Redpesto, I haven't seen any evidence that boys are disadvantaged by single sex education, so what is the cost?

[redpesto]

Redpesto, I haven't seen any evidence that boys are disadvantaged by single sex education, so what is the cost?

But that's the thing: the research is usually into the impact of single-sex education for girls, with a side order of feminism-lite ('those nasty boys won't hog all the lab equipment'), as in this article. As for boys, there is what I'm tempted to call a 'stereotype threat' that all-boys environments turn into 'Lord of the Flies' without a 'civilising' female presence, as in this article. The rest is mathematics...or rather it isn't, because the figures won't add up to make both strategies work.

This, of course, is in relation to state schools: an expensive private education could probably outweigh a lot of gender stereotyping these days. Plus, there's not much of a 'narrative' or incentive for boys to choose 'less sexually stereotypical careers.'

[redpesto]

Gah, the links were missing:

Example 1 (girls schools): http://www.guardian.co.uk/education/2009/mar/18/secondary-schools-girls-gcse-results

Example 2(boy/girl ratios): http://www.guardian.co.uk/education/2007/apr/10/schools.uk

[redpesto]

On the other hand, maybe a boys' school enables boys to enjoy the arts (complete with inevitable picture from The History Boys): http://www.guardian.co.uk/education/2010/jan/20/single-sex-schools-boys-arts