I love science, because the universe has very little tolerance for wishful thinking. You can believe whatever kind of nonsense you like about how the world works, but eventually the data will come along and slap you upside the head. Sadly, not everyone lets the sting of reality affect their prejudices, but that’s another story.
Here’s a fact: among chess grandmasters, there are a lot more men than women. Chess is great, because it’s pretty much a meritocracy, not an old-boys network (colorful parables notwithstanding). There is a simple old-fashioned sexist explanation for this phenomenon, which is that women just aren’t as good at chess as men are. Back in the veldt, you see, when the men were celebrating a successful hunt by playing chess with sticks in the dirt, the women were busy washing the dishes, so there was no evolutionary pressure for them to develop those skills. These days, however, there is a more sophisticated new-fangled sexist explanation for these kinds of discrepancies, which invokes bell curves. It’s not, so the story goes, that the average woman isn’t just as good as the average man, it’s just that their standard deviations are different, so there is underrepresentation at the high end. This hypothesis suffers under the weight of making all sorts of predictions that aren’t true, but it’s kind of scientific-sounding, so it’s gained a measure of popularity in certain circles.
So now someone has looked in detail at the situation in chess. Jake Young at Pure Pedantry points to a study by Chabris and Glickman, “Sex Differences in Intellectual Performance: Analysis of a Large Cohort of Competitive Chess Players.” I noticed the link at Marginal Revolution, and I agree with Tyler Cowen about the most striking findings:
They found no greater variance in men than women. It had been suggested that since science selects for individuals at the upper tail of the distribution, a higher variance in men than women might explain their greater representation. However, the researchers found that — with respect to chess — if anything in most age groups women had a higher variance then men. Upper tail effects do not explain the differences in the numbers of grandmasters…
And:
If you look at the participation rate of women and relate that to performance, you find that in cases where the participation rate of women and men is equal the disparity in ability vanishes. Basically, this means that in zip codes where there are equal numbers of men and women players there is no great disparity between male and female ability — and certainly not a disparity in ability large enough to explain the difference in the numbers of grandmasters.
How about that? It’s not any differences in innate ability, it’s just that women are “choosing” not to play competitive chess. Choosing is put in scare quotes because there’s obviously going to be a great deal of influence from parents encouraging/discouraging their kids at a very young age, but whatever. It’s a shame if young girls who would have been enthusiastic about chess are pushed away by social pressures of one form or another, but for most people chess is not a central part of their lives.
It’s a much bigger deal when women (or whomever) are enthusiastic about choosing something as a career, and are pushed away by an impressive battery of systematic biases. Which is what is clearly going on in science, especially in physics. If girls are given just as much encouragement and opportunity as boys are, and nevertheless choose to become truck drivers or gourmet chefs rather than scientists, that’s fine with me — the goal has never been equal representation of the genders, it’s equal chances for everyone to do what they find interesting. But we have a long way to go before we get there.
Hi Sean,
Great post, Sean. Especially the discussion of how the predictions of the standard-deviation argument are demonstrably false.
It’s depressing how many technically-minded, smart men buy into such facile, demonstrably holey explanations for the under-participation of women in science. Prime example: the letters to the editor in Physics Today, Dec 2006.
Has everyone read these letters? What’s more disturbing: the authors’ lack of intellectual rigor towards hypotheses that fit their prejudices; or that they’re wholly untroubled that their colleagues might think them idiots for voicing such badly reasoned bigotry?
I’ve heard talk of studies (how’s that for citing one’s sources) which seem to show that male brains are more pleased by things such as logical structures and list-making. Could this be a part of the explanation for men’s numerical superiority in chess (and perhaps physics)? Perhaps they are genetically inclined to like such things? I’m not claiming that systematic bias isn’t at work simultaneously.
Here’s an assertion: men are more given to obsessing over things like chess and physics than women. Could this be added to the list of contributing factors?
(and, as a chessnut, I must point out that King Me refers to draughts – checkers to you – but I’m sure you knew that).
Dear Sean,
I only read the abstract from Chabris and Glickman (I don’t have access to the full paper). The findings reported in the abstract are:
a) the male ratings distribution has higher mean, same sigma.
b) matched boys and girls improve and drop out at similar rates.
c) where the gender balance is level, girls and boys get similar intial ratings.
They conclude that the greater number of men at the highest levels in chess can be explained by the greater number of boys who enter chess at the lowest levels.
Leaving aside the fact that this conclusion is almost trivially true, what I think they mean (and this is certainly how this blog interprets it) is that this is the whole explanation – i.e. that the means of male and female grandmaster players in a gender balanced population would be identical.
a) is a statement about the whole distribution including the grandmasters and can explain why most chess grandmasters are men whereas b) and c) refer only to the lower rating levels.
Unless the authors have dealt with this point in the paper, it does not immediately follow from b) + c) that the means for men and women of an *expert* population of chess players with equal gender balance would be the same. Statistics about one tail of the population, interesting and suggestive as they seem, don’t imply that the same is true of the other tail.
Of course you can assume that the difference between men’s and women’s ability to improve is the same at all levels, but that simply puts the conclusion in at the start.
Can someone explain to me the missing logical steps or data – perhaps in the body of the paper?
And I really liked the first 2 sentences of this post. I think many of the world’s ills are down to wishful thinking.
Please don’t bombard me with feminist flames. I am interested in the quality of the argument, not the political correctness of the conclusion.
Yes, but nobody has made any sense out of it all through science and the more I read about Quantum physics and string theory the more it looks like nonsense. Not only that, what you have wrote above is nonsense, because data can’t come along and slap anyone in the face and your first sentence could be said to be wishful thinking!
The more I look at my life and my memories of it, the more I realise that it really does not make any sense at all. Unless I look at Quantum physics and other theoretical physics, then my life is just possible but completely pointless! What’s the point of looking at life like that! Quantum physics and other theoretical physics are more dangerous for your mental health than any other pass time, for one reason, you can work out that you are completely alone and always have been!
Duh! In a society which keeps telling it’s women ‘show us what you’ve got’, it is tough for any woman to go up.
More than the idea that all women are inferior, what is hurting women more are the misplaced goals and ideals.
For all such posts, thank you Sean, CV. I sleep better at night knowing you are speaking out loud. (And I nearly did vomit, and quit physics to be a hermit, after reading those Physics Today letters…).
“What’s more disturbing: the authors’ lack of intellectual rigor towards hypotheses that fit their prejudices; or that they’re wholly untroubled that their colleagues might think them idiots for voicing such badly reasoned bigotry?”
There’s a simpler explanation. Maybe these people are just automatically suspicious of data thrust at them by humorless, sanctimonious, smug lefties. Same goes for global warming: the facts are there, but when they are pushed at you by some sleazy hypocritical lefty oozing self-righteousness [“Isn’t it just OBSCENE that some people want to drive a Porsche??!”], you just want to rush out and personally deforest the Amazon, driving a Rolls to get there.
Here is a simple explanation: Women don’t enjoy math or physics so they don’t go into it. People can be good at something without necessarily enjoying it. I could have probably become a doctor but I hate biology. Also if systematic biases are holding women back in physic are they also holding men back from medicine?
Also its funny that you keep in harping on physics when the number of female undergraduates already outnumbers the number of male undergraduates. Is this because systematic biases are causing men to drop out of university altogether. When is Sean going to start talking about that.
There are many good arguments for looking beyond ‘innate differences’ to explain discrepancies between men and women in various disciplines. However, if we restrict attention to this one issue, we must be wary not to fall into the errors of those we criticize.
For example, in this blog the ‘opponents’ are criticized for cherry-picking data that suggests wider variances in certain skills among men than among women. Yet I’m under the impression that at least some studies do support this result. To ignore these studies because other studies do not give this result is also a form of cherry-picking. It’s not enough to simply find a study that gives a different result; one must be able to explain the previous results as well.
I’m personally inclined to think that it’s very hard to create tests of truly ‘innate’ ability, such that all of these tests are being influenced by social factors and not directly relevant to the question at hand. Perhaps this is a tacit assumption of this blog, and the fact that conflicting results exist is tacitly assumed to be evidence of this.
Yet I don’t think this absolutely settles this issue. For example, there’s an interesting trend that males exhibit autism at a much higher rate than females. Autism is commonly associated with severe social disfunction but I’m under the impression there’s a continuum that might reach well into the eccentric behavior that seems common among scientists (and chess masters?). Now I understand this line of thought is very speculative and tenuous; yet on the other hand autism can be seen as a set of neurological symptoms that occur with different frequency among men than among women, so it leaves open the possibility for others.
I want to close by making clear that I don’t think these academic considerations are the most relevant for understanding the relative participation of women in science or other intellectual pursuits. It seems there is plenty of evidence for social/cultural biases against women, and from a political perspective it is imperative to address these, regardless of any other factors.
A great post! And I think the message is clear. It’s amazing that, even among Cosmic Variance commenters (all of whom are brilliant and enlightened, of course) there is nonetheless some resistance.
I have one minor quibble. As you emphasize:
The scare quotes are essential; they point to the social norms in society which encourage men to play chess (and to do physics), and discourage women from doing so. I think it is these social biases that we’re all striving to overcome. You summarize:
Since “interesting” is clearly societally defined, it would be hard to tell when we’ve accomplished your goal. How would you know whether or not we’ve achieved “equal chances for everyone to do what they find interesting”? It would seem that the only adequate criteria is precisely “equal representation of the genders”. Anything short of this I would take to be an indication of inherent societal bias, and therefore something to strive to overcome.
The comments on this blog on the topic of women in science never cease to amaze (and depress) me.
Pingback: Knitting is a Guy Thing | Cosmic Variance
There might be other physical differences than intelligence or interest that make women less likely to pursue a successful career in science. E.g., a woman who wants to start a family cannot afford to wait until she gets tenure at 40.
Very intriguing topic. There is a conundrum – science vs culinary profession.
‘They conclude that the greater number of men at the highest levels in chess can be explained by the greater number of boys who enter chess at the lowest levels.’
I don’t doubt this is true, but isn’t it odd that there are probably less female professisonal cheffs than male cheffs at the top? Does the societal bias holds here as well, probably an entry barrier or mentoring bias?
The above quote also makes me wonder the future results of the Olympics and other professional sports. As China and India become more affluent, will we begin to see their dominance at the top?
Hope it is not too off topic.
I guess Sean hasn’t considered the possibility of gender itself being an influence on what a person finds interesting on average, or even its influence on the standard deviation of distibutions of how much interest does s/he show for something. Or would that be considered too much of a sexiest statement?
More men than women are in prison.
More men than women get themselves killed attempting stupid stunts.
More men than women become alcoholics or drug addicts.
More men than women are diagnosed with mental illnesses.
And more men than women go into science.
Maybe the “innate difference” is that women are *more* rational. Going into science is fundamentally pretty irrational. Spend years of your life making little to no money, with some chance of never finishing at all, and even if you do, little chance that you’ll be able to find a job in the field afterwards. If you are lucky enough to find a job, chances are you’ll never make enough to justify the years of school. You could’ve gone to med school or law school instead.
It is my impression that a big part of why women are underrepresented in physics isn’t due to what is done, but how it is done. The above mentioned points with competition is probably a factor. If necessary, I can fight, but I’d definitely prefer not to. People wonder why many women drop out at some stage. Yes, the children question is an issue. But more generally than this, I think they just don’t like the way the job looks like in reality, and draw the consequences. There come lots of demands with the job that probably weren’t on your mind when you chose physics as a major. Like, spending an enormous amount of time on administration, proposal writing, advertising your own work. Shaking hands, traveling back and forth and back and forth, moving around the globe, and there is a lot of politics + networking involved. Not everybody is willing to pay this price, and I think women are more consequent to draw the line and say: that’s not how I want it to be, good bye, I have only this life to live. If more of physics were really about physics, I’d think this would happen less.
Mr Knightley: […] men are more given to obsessing over things like chess and physics than women. Could this be added to the list of contributing factors?
If I come to think about chess, well. There were the chess groups at school were all the nerdy guys went, and obviously my maths teacher wanted me to go there. I went there once. See, the reason is simply that I don’t understand how men can be so obsessed about a game. And it’s not only chess. I’ve had a whole bunch of male friends who spent days-and-nights (literally) on some computer game. None of my girlfriends ever did. I had a phase where I’d play computer games, but I easily get bored by virtual reality. Why would I spend my time with a game if there’s a real world out there that wants to be understood?
Maybe I’d be a total looser at chess, but I was much more afraid I might actually be pretty good at it and then spend the rest of my life shifting figures. I too would say there are less women in chess because they just don’t want to. And yes, I find it very possible that gender itself influences what a person finds interesting – on average – a priory that’s got nothing to do with intelligence.
To anon (8):
You say:
Women don’t enjoy math or physics so they don’t go into it.
I’ a woman who very much enjoys math and physics and still have a hard time getting through it– not because of the intellectual challenges, but because of the personal challenges presented by often times being amongst those who imply that I must be inferior due to my gender.
Beyond the ridiculous “you’re a girl, you can’t do physics” remark I received as a freshman undergrad, I’ve had many experiences with less-than-overt discouragement.
Not all women enjoy math or physics, but the ones that do shouldn’t feel as though they’ll need to be 20 times better than their male counterparts in order to succeed. Even with massive amounts of determination, it’s still one of the most difficult things I have to overcome.
The claim that women are “choosing” not to play competitive chess is demonstrably false. I also find the conclusion of this study a bit suspicious. Consider the following statistics from the FIDE (International Chess Federation) January 2007 rating list:
Female to total players ratio in international competitive chess, January 2007.
5029/77144 = 6.5% (all rating levels)
3587/58179= 6.2% (over 2000, superior to strong club players)
1768/39155= 4.5% (over 2100)
697/20743= 3.4% (over 2200)
223/7971= 2.8% (over 2300)
66/2715= 2.4% (over 2400)
10/771= 1.3% (over 2500, grandmaster level)
1/151 = 0.7% (over 2600)
Although men constitute an overwhelming majority (93.5%) of all competitive chess players, there are thousands of active female players on the international rating list. Yet only a handful of them reach the grandmaster level where men are about 5 times more populous than their overall share would suggest. Women can hold up to their 6% representation up to the 2000 rating level or so, but beyond that it falls rapidly for whatever reason.
Its true that men in chess and indeed in physics have a large portion of the ‘grandmaster’ status’s, indeed more than their share of the active already set career population would suggest (an often overlooked discrepancy with the ‘its all environmental biases’ types).
Otoh, the statistics are pretty bad there, particularly in chess. Its quite conceivable a good factor of 2 or 3 of that is just randomness, so its hard to quite make that conclusion.
The bias is pervasive and even the blokes, like Sean, who champion women’s rights fall prey to it from time to time. Here’s an example: we were recently brainstorming on names of possible new folks to join us here at CV. The name of someone I’ll call A Theorist was mentioned. Sean wrote back that A Theorist probably wouldn’t be interested because he’s too high up in the physics political hierarchy. That really struck me. Because I am basically on the same political cmttes/panels that A Theorist is on, plus one more.
I write this as an example as how even folks with the best of intentions can slip and say things without thinking.
It might be helpful to distinguish between the following hypothesese:
SIT (Strong Inherent Theory) – differences in performance (in chess, math, science, whatever) are ALMOST ALL due to differences in inherent ability.
IT (Inherent Theory) – SOME OF the difference in peformance is due to differences in inherent ability.
SET (Strong Environmental Theory) – differences in performance are ALMOST ALL due to differences in environmental factors.
ET (Environmental Theory) – SOME OF the difference in performance is due to differences in environmental factors.
It’s not clear to me whether Sean is arguing against IT, or only against SIT. Arguments in favor of IT are NOT arguments against ET, and vice-versa. It’s only if someone wants to argue for the Strong versions (either SIT or SET) that they have to oppose the contrary view.
My own view is that IT and ET are both true, and so SIT and SET are both false, and that this is compatible with all the evidence offered by both sides in this debate.
The letters in Physics Today are depressing. I have to say that the letters in Physics Today on almost any subject that isn’t the Way It Has Always Been Done seem to draw out the cranky curmudgeons. Like trying to talk about curriculum reform in a faculty meeting. But women in science does tend to bring out the no-reason-to-change brigade most strongly. I tend to think these are related; a field where the curriculum is so constant that you can use a 40 year old mechanics textbook (or 25 y.o. QM) in an first year grad class is a field that feels little pressure to evolve or examine the reasons why it does what it does.
It would be interesting to know what these people think about under-represented minorities in science. Though they invoke unspecified gender differences, I don’t know if they would now be that comfortable in polite society attributing similar differences to ethnicity.
When I was an undergrad, a visiting speaker, in a casual conversation at lunch, told a female undergrad (or said it within her earshot) “Women can’t do physics.” I think few people would feel free to say that today. But the same guy probably still thinks it.
I suspect the reason these threads always degenerate is that nobody treats the
complexity of the problem with the respect it deserves. It’s kind of obvious with
the large array of apologists, but it is also present on the other side too – with
the quota hawks. There are layers upon layers of influences/biases etc both ways. As an illustration, I’d like to point something out regarding JoAnne’s little
dig at Sean.
Committee memberships are not a good measure of seniority, especially in
cases of gender comparison, because women get appointed to committees more
rapidly than men, so the women on high-ranking committees are likely to be
more junior in an absolute sense. I say this because (a) I have female colleagues who perceive/remark on this burden and (b) I have seen the effect
firsthand, when I turned down a committee assignment and suggested a colleague who might serve instead. This colleague has a gender-ambiguous name which is more often used for women than men. The program officer (a
woman) was very enthusiastic until she realised this colleague was, in fact, male.
At that point she lost all interest.
Of course, if everyone adopted positions in the grey area, there wouldn’t be much to post about…
Brad, you are absolutely correct that women get appointed to these cmtte’s earlier on in their career and that they tend to be appointed to more of these things. In fact, it’s something us women have to watch out for and learn how to say no to, unless we feel the work will be worthwhile. However, for the example in hand, the guy and I have essentially the same seniority, we’re only separated by a couple of years, and after just checking I found that I actually have more publications and cites.
Most importantly, please let me be clear: I was not making a dig at Sean. There isn’t anybody I respect more. I was merely trying to illustrate how easy it is for any of us to say something that can discount a woman’s impact. It truly happens quite often! I apologize to Sean that my comment was so poorly written and so easily misconstrued.
Now, back to my calculations of Supersymmetry at the Linear Collider…