Monday, March 22, 2010

Kinda Proving The Point

NYTimes:
At the top level of math abilities, where boys are overrepresented, the report found that the gender gap is rapidly shrinking. Among mathematically precocious youth — sixth and seventh graders who score more than 700 on the math SAT — 30 years ago boys outnumbered girls 13 to 1, but only about 3 to 1 now.

“That’s not biology at play, it doesn’t change so fast,” Ms. Hill said.

Another hypothesis:
The test has gotten easier, or more people know how to game it, and consequently more children are scoring over 700. From this chart it appears the average math score has indeed risen over the last 30 years or so. Presumably, more are also scoring over 700. Larry Summers' theory was that there are more boys in the tails of the distribution, that is to say the high and low ends. While women have similar math skills on average, they have a smaller variance around the mean. If we shift the distributions, toward the higher end, then it is entirely possible that women "catch up" without actually catching up.

Basically, what I am saying is that getting over the 700 barrier has gotten easier and that because of the statistical distributions it would shift the proportion of females-to-males.

Note: When you are writing a piece about how women don't suck at math and the problem is bias, try to make sure your statistical analysis is tight.

This, however, is a pretty good point:
Even if male math geniuses outnumbered female geniuses 3 to 1, Dr. Hopkins said, it would be reasonable to expect one female math professor for every three male professors at places like Harvard and M.I.T. “But in fact, Harvard just tenured its first female, after 375 years,” said Dr. Hopkins, who, famously, walked out of the room after Mr. Summers made his controversial remarks.

It certainly proves there has been historical bias over the previous four centuries, which is something I think nobody would dispute.

Then there's this:
The university women’s report cited research showing that girls’ performance suffers from any suggestion that they do poorly at math.

Isn't this the fragile feminine psyche defense?

If you asked my straight up whether I think there is bias in math and science, then I'd say straight up - yes. But when the statistical analysis is iffy I get pissed off.

Oh, and in case you are wondering, Miss FLG will not get away with any bullshit about girls don't do math. She's learning calculus even if it kills her.

6 comments:

The Ancient said...

Even if male math geniuses outnumbered female geniuses 3 to 1, Dr. Hopkins said, it would be reasonable to expect one female math professor for every three male professors at places like Harvard and M.I.T.

She lost me at "reasonable."

Andrew Stevens said...

Prior to 1995, the average combined SAT score was about 880. It was recentered in 1995 so the test would again have an average score of 1000. This dramatically raised the number of people who score 700 or higher. Prior to 1995, a few dozen people a year attained a perfect 1600 on the SAT. After 1995, that figure is now in the thousands every year. (Note that I don't know what the effect of adding a third section and raising the top score to 2400 is. Too recent for me to have bothered to look into.)

So the short answer is that your theory is clearly plausible and Ms. Hill's theory is explaining a phenomenon which does not exist and does not need to be explained.

Anonymous said...

"She's learning calculus even if it kills her."

You haven't had an adolescent in your house yet... She's doing what she damn well wants, and be grateful if she doesn't make you a grampa in 13 years. And by the way, you may want to lobby to get Virginia to boost penalties for folks who do piercings for under-18s without their parents' consent. dave.s.

Flavia said...

I'm certainly not arguing that all differences in performance between men and women (whether in science/math fields or anywhere else) can be chalked up to social factors. But I also don't think that the argument that the perception of female intellectual inferiority affects women's classroom/test-taking performance amounts to an argument about the fragility of the female psyche; haven't there been a number of studies that have shown that the performance of both girls and racial minorities on certain kinds of standardized tests have been stronger/weaker depending upon the message those groups have been given beforehand about the "natural" strengths or weaknesses of their sex/race?

Obviously, I'm no authority on the subject (and have not yet had the chance to look at the full, 134-page AAUW report), but I know a lot of people who work on childhood development/school achievement and the like. There's also an interesting discussion of the effects of beliefs about intelligence on performance (as well as lots of stories about workplace discrimination) at Historiann's.

Anonymous said...

My wife is the daughter of a physicist, who had Views that his kids would damn well learn calculus - rather similar to yours. She tried, with good will, it didn't work. His son, also, had a far more literary bent, and the attempt to cram maths down his throat yielded alienation for years (well, along with other things). My suggestion here is that it's worth trying, and that you don't bet the ranch on it. Some kids are math people, some aren't. dave.s.

FLG said...

Dave.s.:

I more worried about the piercing thing now.

 
Creative Commons License
This work is licensed under a Creative Commons Attribution-No Derivative Works 3.0 United States License.