New study on gender gap in mathematics/reading

[Aneo.]

Go to any math or science camp or competition, and it is no secret that boys are overrepresented in these subjects. While we have long thought that men where more "spatially" oriented and hence better at math (and thus poorer at reading), with even the president of Harvard having to resign over such a controversy, a new study suggests that the gender gap is much more cultural than innate:

http://education.guardian.co.uk/schools/story/0,,2283083,00.html

Discuss.

(Hopefully we can have an intelligent discussion on this here, but if not, move to GFF)

[guineapiggies]

hmmmm that's an interesting study.

i took some online classes for math and english. the math classes r dominated by boys the most girls in a math class that i have been in is 2. me being one of them as for my english class. we have 4 girls and no boys.

our school math club had about 10 boys and 6 girls.

[MysticTerminator]

for comparison's sake, the IMO participants of the more "gender-equal societies" of Iceland, Sweden, and Norway (this is based on my best guess -- some of the pics are, well, hard to tell)

2008

Iceland 5 M, 1 F
Sweden 5 M, 1 F
Norway 6 M, 0 F

2007

Iceland 5 M, 1 F
Sweden 6 M, 0 F (I think?)
Norway 5 M, 1 F (blargh hard to tell wtf)



also for comparison's sake, the composition of our own team (in a country which does have a gender gap):

2008 6 M, 0 F (probably, haven't kept careful tabs)
2007 5 M, 1 F
2006 6 M, 0 F

the same trend would probably continue for Putnam and IMC but it's a bit harder to separate based on country there, so I won't bother looking up that data

[jmerry]

The historical fact that these gaps have always decreased over time points to an origin in culture and discrimination. Consider the US IMO team: no girls in the first 24 years, three appearing a total of five times in the 11 years since. The gap is still there, but going from none in 100 to about 5-10% is a big change.

The Putnam wouldn't be useful for international comparison here, since it's USA and Canada only.

Some AMC statistics (2008):


USAMO data isn't easily available; only names are available on the site. The gender data above is self-reported, and the substantial fraction who didn't report anything are not included.

What does the gap look like in the contest world? There's a participation gap, with significantly more males than females showing up at the open contests. This gap grows with age; it's larger for the AMC 12 than the AMC 10. There's also an achievement gap among those who do participate- about 9 AMC points. This gap appears at all levels of success, leading to much larger gaps in the participation numbers for selective contests.

[MysticTerminator]

Minor point -- many good students come from overseas to compete in the Putnam including Chinese IMOers (who always take up a couple of the fellow spots), Russians, Romanians, Moldovans, and so forth, which is why it may have some (but admittedly not a large amount of) use even in international comparison. That's why I noted the IMC, though, since yeah, American data is probably not terribly interesting as we're already pretty aware of that.

[Aneo.]

I think the study was more related to the average student, not IMO participants. There might still be a social stigma against girls in these "gender equal" countries about math competitions (though perhaps not, I'm just hypothesizing). Also, there have already been studies that try to explain why in "very high level" activities, males are overrepresented.

http://www.dailybruin.ucla.edu/news/2005/jan/28/general-intelligence-equal-bet/

If I recall correctly, another study (or perhaps this same one) described that the average intelligence of males and females were exactly the same, but females had a lesser deviation from the average, whereas in males there was slightly more deviation (i.e. more smart and dumb people).

Of course, there are those that say IQ is a useless measure anyway, so the study might be completely moot. I think, however, that is irrelevant.

But again, I think the study was geared to the basic, average student, and I think the study was interesting when considering this.

Also, note that countries where there is a wide social stigma against women (e.g. Middle East and Asia), female participation is virtually nonexistent, compared to those in more "gender" equal countries (e.g. Northern and Western Europe), where there is a noticeable female participation, if still underrepresented.

[fedja]

[jmerry]

Some innate differences are plausible- but they're tiny compared to everything else. Success in problem solving starts with liking math, and the "Math is hard/You're no good at it" message drives people away.

Continuing my side of the thread- the gap in contests:
Also, at the elite level, the gap is shrinking. Compare MOP now to ten years ago: in my years (1998, 1999), Melanie Wood was the only girl both times (out of 24 and 24). This year, there were at least 9 (Colleen Lee, Wendy Mu, Lynnelle Ye, Carolyn Kim, Jennifer Iglesias, Katherine Rudolph, Elizabeth Synge, Shijie Zheng, Iris Xu) out of 56, and maybe some I missed with non-English names. That's still underrepresented, but it's a big change in only ten years.
One striking pattern in those nine names: they don't come alone. Six of them come from schools with two or more MOPpers, including two pairs of two girls from the same school. It's also worth noting that Melanie Wood went to the same schools as Chris Mihelich, who reached MOP in 1997; she wasn't alone either.

That's the gap we see now: three or four times as many boys as girls taking contests and problem solving seriously. It's closer to equality where the contest culture is strongest, and farther away where only a few interested kids stick out.

[MysticTerminator]

whoa there, jmerry, that's simply due to the invitation of the CGMO participants to MOP. Not to mention the fact that red MOP has since been added. If we look simply at blue/black MOP during the past years, the number fluctuates between zero and two, I believe.

[jmerry]

OK, I didn't realize that difference. The 12th graders involved should have been a clue.
As for red MOP, I'd prefer to include it, and simply compare to the total size. That's not the only change anyway- there was no TST before 2000.

My other point stands.

[SimonM]

Perhaps taking the MOs as an example is a bad idea. Perhaps it just shows that boys are naturally more competitive

[fedja]

Well, you can then argue that looking at the performance on the standard tests is even a worse idea, because, perhaps, it just shows that girls are more disciplined, more inclined to do what the teachers tell them to do, and are more inclined to do their homework properly (this is rather undeniable and nothing else is really required to prepare well for a standard test, while just being more competitive would not bring you far as long as MO is concerned).

The last remark is sort of pointless but once the Harvard president had to resign for just openly stating his opinion (which, in my humble opinion is a shame to the society that takes pride in its free speach rights), I do not think there will be anybody in the USA in the near future who will dare to argue that boys are any better and the studies with the conclusions like the one I quoted in my previous post will appear like mushrooms after a rain.

[JBL]

I didn't notice this in the O.P.'s post, so I'm going to quote Fedja instead:

[echen]

Girls just dont seem pressured to do things like math and science. In my advanced math class their is one girl in the 22 students.... Also, people tend to do things their friends like to do. If math is all boys, girls wont want to participate and so math competitions are dominated by boys.

[dreambold22]

i agree with that. also, if they think they CAN'T do it, or they dont feel it's important, they won't try at all. i learned this from my MATHCOUNTS 6th-grade group at my school: all the girls said that they didnt really care what they got on the sprint, they didnt have a chance. then, when the 6th grade top 5 list came out, i was the only girl on it (5th place ). probably 'cause i was the only one who actually TRIED. but i still sucked at it (10 out of 30).

[Kent Merryfield]

I'm just an observer of local conditions - I don't have any answers. I teach at a non-doctoral state university, and I coach the Putnam team at my university and the ARML team for our region. I'm particularly interested in the question of trends over the last few years - which is what jmerry has been talking about.

Our university admits students based on a strict numerical formula which involves grades and test scores, and weights grades quite heavily. The result is a freshman class which is about 2/3 female, since girls, on average, make better HS grades than boys. The math major is pretty close to 50/50. That is more male than the university as a whole, but it's still a lot of female math majors. Of course, a significant fraction of our math majors are planning to go into high school teaching, and teaching is a traditionally female profession. If I ask who the best 5 or 6 graduating seniors in the major are, some years it seems that they're nearly all male and some years nearly all female - it's hard to spot a trend.

However, the engineering students and my university are overwhelmingly male, and if anything, have become even more overwhelmingly male in the last 5 years or so.

The handful of students I talk into risking their egos by taking the Putnam exam tends to be about 70% or 80% male. And in most years, the three-member Putnam team from my school has been all-male.

As for the ARML team: I've been doing it since 2002, so I have 7 years of records. Counting just people who live in what is now our territory, I've got 219 team members (counting the same person more than once if they were team members in alternate years. Of those, 54 are female: 20%. We have a multiple-team organization. Who has been members of either the A team or B team, the more carefully selected teams? In that case, 29 of 154 have been female, 15%. But, and here's the odd part: the percentage of females on the ARML teams has been going down.

The overall percentage was above 20% in 2002, 2003, 2005, and 2006. It was below 20% in 2004, 2007, and 2008. For the A and B teams, the high in female membership was 7 out of 30 in 2003; the low was 1 out of 30 in 2008.

I don't really know why that has been going down.

Of course, the ethnic mix of the ARML team does not come remotely close to matching the ethnic mix of the metropolitan area. The ARML teams have been 75% Asian, including nearly 50% Chinese. Of the 20% or so who were white nearly half were either Jewish, immigrants from Eastern Europe, or both - and that's not exactly the overall demographics of the metropolitan area. In the face of numbers like that, which must have a cultural explanation, one's first thought about explaining gender trends leans towards the cultural (although it is true that the girls on the ARML teams go to the same schools and in some cases belong to the same families as the boys on the team.) But that's as far as I'll go. I didn't promise any explanations.

[durt]

[HilbertThm90]

Um...How do you test reading? I tend to look quite suspiciously upon any research that claims that have a way to score reading. If they are talking about ridiculous tests such as: read this paragraph, which of the following could be the title of the paragraph, then I don't buy it.

I say the actual result is that females are better at answering absurd test questions that have little to do with how well you read. I do horribly on those tests, but I am a fantastic reader. I've read huge major works (Underworld, Infinite Jest, Gravity's Rainbow, The Brothers Karamazov, etc) and had great discussions about them.

Sorry. That is a sort of sore subject for me doing poorly on the reading of the SAT and GRE due to questions of that sort. Maybe I picked the title that best fit because I realized a subtle irony in the piece and thought it should have an ironic title to draw that out (even though the answer is some boring obvious choice). Those questions are not good!

[Dojo]

I've also read somewhere that while girls may be better at speech, that makes them slower and therefore enjoy less math. This is because when girls solve a problem, they spend too much time putting everything into words, making them slower. When anybody feels slow, they may be less motivated.

[Quickster94]

hmmm, this is very interesting. I wonder if we had data about a culture gap? In mathematics competitions, it seems east asian and indian ppl make up a unnaturally high percentage, and hispanics and blacks unnaturally low.