Many of you may have read about the recent study, "Gender Similarities Characterize Math Performance," which was published in the July 25 edition of Science; the results were fairly well publicized in the popular media. (I'm sorry I can't link the Science article, but I have not found an ungated version of it.) The research attempts to determine if there gender differences in math performance that could help explain the relative lack of women in STEM (science, technology, engineering, and mathematics) careers.
In the study, standardized math test scores in 10 states, covering students in grades 2 - 11, were examined for gender differences. The primary finding reported in the media was that there was no significant difference in the mean math scores of boys and girls.
The researchers also compared male and female performance on hard problems requiring a higher level of complex problem-solving found only a slight difference, with boys scoring a bit higher than girls at grade 12 (d = 0.08). However, they had to use a different data set because the other tests did not include any difficult problems. Even the second data set only had problems with difficulty of 3 on a scale of 1 - 4. So it's arguable that there is no evidence here for how well students would do with the most challenging kinds of problems.
One big finding that was either not reported at all, or misleadingly reported, in the popular media accounts that I read concerns the variability in math performance. In this study (as in others I have read about before - the Science article says this hypothesis has been around for over 100 years), they found that despite having equal average scores, there was a greater amount of variability in boys' scores than in girls'. This means that boys will be over-represented among both high and low scoring students. Looking at the scores of white students, at the 95th percentile, there are 1.45 times as many boys as girls, and at the 99th percentile, there are 2.06 times as many boys as girls. The researchers say: "If a particular specialty required mathematical skills at the 99th percentile, and the gender ratio is 2.0, we would expect 67% men in the occupation and 33% women."
I think there is a disconnect between the views of the general population and academics on this issue. The idea that girls are worse at math than boys, on average, is a belief that endures among the masses, and I'm certainly in favor of the press refuting the "girls as math morons" stereotype with the latest scientific findings.
However, academics and their fellow-travelers (those with knowledge of the literature in this field) have, for the most part, been convinced of this already. The academic argument has long since moved on to the variability issue. And when you are discussing the very high level of math performance required to succeed in STEM PhD programs, 99th percentile is not necessarily high enough, in my opinion. If math ability is normally distributed, then the farther you get into the tail of this distribution, the more boys will be advantaged.
Of course, there is another disconnect between those who are primarily concerned with scientific research of this kind and those who are motivated by a strong desire to see more women succeeding in all levels of STEM occuptions. It is not mere paranoia for the latter group to be concerned that any discovery that boys are more likely to have superior math ability will be latched upon as a justification of the lack of women in STEM positions, regardless of what other evidence exists. (If you look at the comments section of this blog post, for example, you will see a some triumphant dick-waving among the male commentariat.)
All the same, I have been unimpressed with the pressure brought to bear on people like Larry Summers to pretend that ability differences, preferences, and other non-discrimination-oriented factors cannot possibly be partial causes. Have any of you actually read the transcript of the infamous speech that he gave on this subject? If not, you might be enlightened by at least scanning it. For something that has been portrayed as one step more subtle than him thumping his chest and yelling "Man, Smarter Than Woman, Ugh!", it's actually pretty nuanced. It is almost unbelievable to me that a woman with her shit together enough to herself be a professor in a STEM field would be brought to hysterical, angry sobbing upon hearing these words, though that was reported by the media. (Note: I am using the term "hysterical" advisedly.) Maybe people weren't listening when he said, "The other prefatory comment that I would make is that I am going to, until most of the way through, attempt to adopt an entirely positive, rather than normative approach, and just try to think about and offer some hypotheses as to why we observe what we observe without seeing this through the kind of judgmental tendency that inevitably is connected with all our common goals of equality."
As for me, I don't have any ready made answers to this problem. I think all of the things Summers mentioned, including outright sex discrimination, are factors that influence the under-representation of women in the STEM professions. Even if we had an overwhelming amount of evidence that women are less represented amongst the elite math brains, math ability is only one necessity for becoming an engineering professor at MIT. And I think it is our responsibility both to continue rigorous scientific inquiry into the causes (including the free dissemination and debate of this research) and to encourage girls/women to enter these fields and fight discrimination wherever it is found. It's okay to treat the issue as both an empirical and a moral question.
Subscribe to:
Post Comments (Atom)
5 comments:
There are people (esp. in academia) who think that the 'variance' issue is also a cultural thing, a result of discrimination, and so stating it as a 'positive' fact is, in their view, a normative position. Or believe that there is no 'positive' and 'normative' distinction - that all facts which are identified reflect the biases of the observer. To a great extent, they led the witch-hunt against Summers.
(These are the same people who think evolutionary psychology is just an evil tool of the patriarchy used to justify the status quo of male/female relationships, not a valid scientific technique based on the basic fact that humans are just another species of animal, and gender differences are seen in physical and neural structures of male and female animals in myriad species, and we have some idea of what lead to their evolution.)
How do you separate potential physical differences that may dictate some kind of prescriptive capacity for math from the way boys vs. girls are raised and what expectations are set, etc? I'd also be curious how the girls rate themselves vs. boys in terms of their math/problem solving skills? I.e., for the same score, is there any difference in girls thinking they're worse at math than they really are (esp. if girls believe that there are fewer girls who excel at math than boys)? Or any difference in how good at math/science they think they need to be to pursue a career in that field (or even how much they know about their career options)? To Robert's point, we could also ask, of women who are good at math, how many choose not to have children versus men?
And of course, there's the problem that a man's choice to have children does not have the same career-disrupting consequences that a woman's choice does. That fundamental biological disparity has to be a big part of this.
These are all good questions, but I am particularly interested in the one regarding how good a boy vs. girl thinks he/she needs to be at math to succeed in a STEM profession, and in the related question: how good does a boy actually need to be in math versus a girl? Girls may think they need to be better, and they may be right, due to discriminatory factors.
Speaking of discrimination and stereotypes... happened to have a male IEEE employee (who is trying to get our business by the way, in the form of a corporate IEEE subscription to publications & standards...) ask me if I went to college. Blown away by my response, it apparently changed his entire perception of me in a way that previously telling him I was an engineer on a technology R&D team simply didn't do -- he still thought I was "administrative" until realizing that a math/computer science degree from Stanford would be a little overkill for that. I'm still fuming.
Jesus, Jen - that is so infuriating!
Post a Comment