"There's nothing like a friendly reminder of the creepy power of stereotypes and suggestion to put the stakes -- a girl's or woman's personhood, really -- in perspective."

That works for boys and men, too, directly (do men more often define themselves as math wonks or artsy? One guess) and indirectly in how they engage with girls and women.

Scary stuff!

Crazy stuff! I think it also shows a reason to be against high-stakes testing. Are some schools not meeting No Child Left Behind standards because of the subtle influences of a flag or picture on their classroom wall?

FYI: The Washington Post article said an American flag on the wall boosted the scores of white students, but had not effect of minorities' scores.

So many of these cues, you wouldn't even notice conciously. We can't control them all, but it's good to be aware. And perhaps not judge achievement so strongly by "the numbers."

It's been known for years that how questions were asked, in what order, grouped how, and how worded affects the results of a survey. I would have to see this survey and the steps the authors took to interpret it before getting too worked up about the announced result.

At face value, it re-proves that the order of questions and their subjects alters the response. It is a huge scientific leap to say that their interest in art or math was actually altered. I got double degrees in Chinese lang&lit and Physics. Which do you think I said my major was when I met a girl in an average bar?

People's minds and interests are pretty resilient and will easily survive a questionaire. They may not survive constant browbeating or public humiliation, the two most common causes of math-phobia (fear of math, the one with the sweats and panic attacks -- not to be confused with being worried someone will think you're a nerd).

Girls: Be Architects!

Then, when people ask you what you do, you can say, "I'm an Architect," and you either can glory in the artistic Beaux Arts idea of "Frozen Music," or you can geek out on the technical nuances of wind-sheer loading ratios.

And the people who you are talking to can pick whichever side they prefer. You can't fail to impress, and you need never define yourself, even to yourself.

As a matter of fact, now that I see this written out, I recommend that you tell people you are an Architect, no matter what you really do for a living.

Why is that so disturbing....isn't that right up a psychologist's alley? Lightweight.

the concerted portrayal of men over the past 20 years as bestial, stupid and deserving of physical abuse is such a bad idea. Not that feminists would notice or care, given that they started the whole thing, but if women can be so powerfully influenced by even the most subtle suggestions, how are men affected by the pervasive misandry in popular culture, academia and elite liberal discourse?

Is this really about art or math preferences? Or are the cues telling us that women are easily tuned-in to hint and innuendo? I would also like to read the questions that were asked. Also, could it be that some women like both but are a bit ambivalent and so pick up questioners cues and lean one way or the other according to the question's bias?

and I'm not surprized at all to learn that specific words influence response patterns in either women or men. It's a fascinating, amazing aspect of humans that our complex brains are pre-built with reflexes for group behavior... survival for us relates to belonging, leading, following, etc., and so much of our behavior depends on our perceptions and surroundings. (It's certainly a well-known fact to folks like Procter & Gamble and Josef Goebbels.)

I'm not sure this set of reflexes indicates deep-set interest in math vs. the arts, but I am sure it's inspiring to a whole new generation of mind benders!

(I hate the ad industry by the way, it gives me the creeps... get me outta here!)

This basic concept of the self-fulfilling sterotype has been known to be true for a long time in both males and females, replicated by many psychologists using varied approaches. Steele (Claude, not Jennifer) and Aronson described it in 1995 as "stereotype threat." For example, they showed, that if one asks about race or ethnicity before a standardized test, African-Americans perform worse on average than they do when they take the same test but are not asked about race first (the stereotype in question being that blacks score lower on such tests). The same is true for women asked about gender before a math test, as well as several other race/ethnicity/gender stereotype situations. It's interesting to note that this doesn't always cause worse performance. In at least one study, when Asian-Americans were asked about race/ethnicity first, they scored higher on a math exam than if the race question was not asked. I suppose the contribution of this new study may be the subtlety of the cues that you can use to get the stereotype threat phenomenon, but the basic idea that when you are reminded of your stereotyped-group-member status, you behave more like the stereotype predicts is very well established.