Reader Tracy Allen sends some interesting information from the College Board’s report of recent test results on the SAT 1 of college bound seniors.
Our friends at Census say that in the 15-19 age group, men make up 51.4% of the total population. But they only represent 46.5% of SAT 1 test takers. I’ll be the first to admit that there are lots of things could explain that away….
But the score distributions show that among the test takers in the higest score range (750-800), males made up 50.5% of that group on the verbal reasoning portion of the test and 68.7% of that group on the math reasoning portion of the test. In fact, males are disproportionately represented in the score ranges on the math section for all score ranges above 600. Looking at the most extreme tail of the distribution, the coveted 800, males were 49.3% of the population achieving that score on the verbal portion, and 72% of the population achieving that score on the math portion. So, even in the verbal reasoning portion, men are represented in a higher percentage than they are in the overall test population. Without question, though, they are dominant in the math portion.
Tracy concludes, with good reason:
This is remarkably consistent with Larry Summers heresy, only here it’s being perpetrated by the College Board. I guess they’ll be in the cross hairs soon as well.
I found surprising the section on intended college majors, as it showed students to already be assessing well their own strengths and interests. Those intending to study Engineering, Mathematics, and the Physical Sciences had the highest scores on the math section (and broke 78%/22% in favor of males, to keep this on-topic). Those intending to study Language and Literature; Foreign or Classical Languages; and Library and Archival Sciences had the highest verbal scores (and broke 71%/29% in favor of females). Interestingly, those intending to pursue the Physical Sciences (is this not what your daughter is studying, John?) had the overall most impressive scores amongst intended majors, ranking 2nd on math, 4th on verbal, and 2nd for total score.
Fancy that — those wanting to do math-related stuff are best at math, and those wanting to do language-related stuff are best at language.
I’m curious about the people in the intersection of the two 750-800 groups (I was in that group way back in 1991)… what’s the composition of that group?
Feminists with Vapors
Via John Rosenberg at Discriminations, this bit of statistical information about women, men and SAT scores.Our friends at Census say that in the 15-19 age group, men make up 51.4% of the total population. But they only represent 46.5% of
“This is remarkably consistent with Larry Summers heresy,”
That men are good at standardized tests?
Nels – Yes, Jessie is studying Applied Physics. I hadn’t even realized there was such a thing as applied physics, much less departments of it, until she applied to grad. schools. Not surprisingly, she did quite will on all the standardized math tests, but somewhat surprisingly she did slightly better on the verbal tests. For what it’s worth (see, I’m trying to keep on topic, too), approximately 7% of physics professors are women. (If her present plans hold, Jessie wil not raise that number, as she does not now intend to go into teaching, at least initially.)
Gosh I wish I could come up with a test- any kind of test, physical, mental, combination- in which scores were distributed to match diversity categories as they occur in the “general population”. No one else seems to be able to.
I’d get so rich…
“That men are good at standardized tests?” That would be an article of the Larry Summers Heresy, no? :-)
So who is it that schools are shortchanging, today?
Laura –
You raise a good question. One of the interesting, and perhaps counterintuitive, elements of the College Board Report, is that relative to males, females reported a significantly higher mean GPA (3.35 to 3.19), and a higher percentage in the top tenth in class rank (58 to 42).
Math was the only subject area where the male mean GPA was higher than female (3.08 to 3.07). Additionally, females were equal or greater to males in the number of years of academic study for all the subject areas.
So according to standard school measures of performance, females are doing better. That is what makes the SAT I outcomes all the more troubling. In fact, the report lists 33 years worth of mean SAT/SAT I scores for College Bound Seniors. On the verbal portion the male mean has ranged 2-13 points higher than the female mean while on the math portion the male mean has ranged 34-42 points higher than the female mean. And while the means differ among various ethnic groups, these ranges in scores are remarkably consistent within a given ethnic group, so this really does appear to be a gender based issue.
Certainly some of this can be explained away by the self selection of the population – fewer males who would score low/lower sit for the exam. But that is not sufficient to explain away the substantial difference in scores on the math portion of the exam.
My concern with the whole Harvard/Summers dust up is that the very people who should be researching differences like this and looking for substantive answers that contribute both to our understanding and our ability as a society to address it constructively seem all to eager to shoot on sight anyone who dares to question the prevailing orthodoxy, which seems as grounded in dogma as it is in either history or science.
” That would be an article of the Larry Summers Heresy, no? :-)”
I thought his heresy was men are good at science, which is different than being good at standardized tests.
There’s a simple, obvious reason why women have higher college GPAs, even if men have higher SAT scores:
Men are more likely to have academically challenging majors, like engineering, physics, chemistry, biology, math, etc. – while women are overrepresented in easier majors like education, art, etc.
Let’s compare the grades of men v. women in EACH class – i.e, in calculus 3, advanced biology, etc – to get an apples v. apples comparasin. College GPAs v. SAT scores is an apples v. oranges comparasin.
REVISION of previous post…
Actually, I should say that the only way to accurately compare college GPA to SAT scores would be to force an equal percentage of men and women to major in each subject (engineering, math, physics, education, etc), and THEN compare the GPAs. Comparing grades of men v. women currently in engineering, physics, math, classes is not an accurate comparasin with SAT scores, since you’re getting such a small number of women in the sample, who are so far above the average math level of the average woman in college.
The following working paper estimates a regression of student grades on student characteristics. Controlling for student SAT scores and high school rank, and also for the subject and level of the course, male students still perform significantly worse than female ones.
http://www.clemson.edu/~hrey/hernandez_gradeinflation.pdf
The following working paper estimates a regression of student grades on student characteristics. Controlling for student SAT scores and high school rank, and also for the subject and level of the course, male students still perform significantly worse than female ones.
http://www.clemson.edu/~hrey/hernandez_gradeinflation.pdf
As I stated in another thread, if you’re at all interested in the broader topic of gender issues in education, pick up a copy of Chritina Hoff Summers’ “The War Against Boys.” While it examines many issues in great detail, the book basically notes that modern schooling, from primary through college and beyond, is biased toward women. That could help explain why there is a difference between performance on a test to measure apptitude (which may have committed the sin of not being biased toward females), and performance in a classroom setting, which, as Hoff Summers notes, is biased against men.
To clarify the above post, education is biased in favor of women. The word toward can be a little confusing there.
If education is biased against boys, how is it that boys’ SAT scores are higher? SAT is supposed to be a predictor of school performance, not pure g.
Re the Clemson study:
Really, the .124 difference in GPA is not that high. Also, it doesn’t neccesarily give more weight to more difficult majors. It gives more weight to majors with a tougher average grading curve, which is totally different. For all we know, male-dominant majors like math, physics, engineering, etc. may have higher average GPAs than female-dominated majors like education, art, etc. So possibly, even though male-dominated majors may be tougher, maybe they are actually weighted less in the study. We just don’t know. Also, even IF the grading curve is a little tougher for male-dominated majors, it’s pretty clear that physics, engineering, etc, takes a lot – not just a little – more effort and intelligence than stuff like education, art, etc.
Bottom line: In college, men and women take vastly different classes in different proportions (Men on average have much tougher classes, particularily in math-related classes). Therefore, comparasins are less valid than in high school, where there is little variation in class choice.
One reason why men are overrepresented at the high ends of the Bell curve for the SAT (the highest scorers) is probably because there is a higher intellegence standard deviation among men. There are more really smart and more really dumb men. The smart men are math, engineering, etc majors. The dumb men aren’t in college (why fewer men than women are in college).
So women are basically equal to men at math ability, since their college math GPAs are basically equal (3.08 vs 3.07 GPAs in math classes)?
Uh…OK…so if women get a “B” in a General Education lower division math class (Algebra or Stats) that’s the same level of an accomplishment as getting a “B” in each of the several calculus classes that are required for male-dominated majors in math, science, and engineering? How many people would really believe this???
My daughter is a junior in high school and just received an 800 on her math portion of the SAT exam she took in March ’05. She is an excellent physics student. What does the future look like for women in fields pertaining to physics and math?
Our daughter (a junior in high school) just received an 800 on the math section of her first SAT. She is a really strong physics student. But what does the future look like for woman in fields pertaining to math and science? Has Larry Summers set women back 50 years?
The reason boys do better on the SATs (and GREs) is that no one is discriminating against them very much there.
Anyone who has taken the GRE will tell you that the Quantitative section (the ones boys are supposed to be better at) is very easy, whereas the Verbal section is made substantially more difficult. So the system is inherently biased against boys there too. But boys still do better on both sections. Maybe it is because years of being given good grades without having to work for them has seriously screwed girls up mentally. :)