Thursday, May 13, 2010

Are men smarter than women? Your guide to detecting bad science

The UK's Daily Mail recently published an opinion piece by Richard Lynn stating "on average, men are more intelligent than women."  Moreover, Lynn asserts that at the near-genius level of IQ, "brilliant men outnumber brilliant women by 8 to one."
In this context...indignation that only one in ten science professors is female doesn't seem all that bad. It also goes some way to explaining why, in almost 110 years of Nobel Prize history, only two women have ever won the Prize for physics, only four have won the Prize for chemistry and why no women at all have ever won the coveted Fields Medal for mathematics in eight decades of trying.
Making this argument is pretty much exactly as ludicrous as saying the fact that no women were recorded to have voted in federal elections before 1920 shows us that women are, on average, less politically inclined than men.  And Lynn's article is chock full of such gems.

When I expressed my horror on twitter, a friend wrote back saying "It's the Daily Mail!  And it's Richard Lynn!"  In other words, don't waste your time.  Yes, this is Richard Lynn, staunch advocate of The Bell Curve, eugenics, and other racist pseudo-science.  And yes, the Daily Mail isn't exactly the Wall Street Journal.  But, sorry, I just can't bring myself to sit this one out.  Because the fact that Richard Lynn can still call himself a scientist, and his theories and the evidence he provides to back them up can be thought of as scientific, is an embarassment to every real scientist out there, and to those of us working our tails off to put the letters P, H, and D after our names.

Jezebel rightly points out that people make arguments such as "X group is just naturally better at A, B or C" to avoid addressing social inequalities that would be painful to remove.  But the fact that those social inequalities exist make it impossible to answer a question about whether men or women are naturally more or less intelligent.  Characteristics such as sex and race are immutable, and inseperable from their underlying genetic identifiers, meaning there's no way to observe the effect of having a "female" or "black" genotype without having the accompanying socialization that people with those characteristics experience.  (Yes, there are phenomena such as "passing", race is socially constructed, people can choose to associate with one race or the other, and individuals can choose a different gender than the one their sex characteristics assign, or, for that matter, be born intersex.  But, by and large, individuals with female sexual characteristics will be treated as female, at least for certain formative years, and people with black skin will be treated as black.)  So if, for example, female children are encouraged to play with dolls and not fire, on average, we cannot separate the effect of this female socialization from the genetic makeup of femaleness.  Because of this, no scientific data can answer the question Lynn claims he knows the answer to.  If you had an unlimited budget to run a randomized experiment, you could not design one that would separate the impact of genetics on intelligence versus the impact of society's reaction to the observable pieces of those genetics.  Lynn's "scientific hypothesis" is really a Fundamentally Unidentified Question.  In other words, FUQed.

When the Harvard law racist email scandal broke, Jesse Taylor over at Pandagon noted that the fact that people were always asking whether black individuals are genetically less intelligent than white individuals, but never the reverse, revealed that their true motivation was racism, not scientific inquiry.  But I'd go one step further: Anyone who asks any question of the form "Is X group more [socially constructed characteristics] than Y" is not doing so out of intellectual curiosity.  It doesn't matter if we're asking "Are black individuals smarter than white individuals" or the opposite--it's still a bad question, and it's still not science.   Not just for the reason above, but also because the concept of intelligence is itself socially defined, and according to a standard which was designed by the very white males who seem to somehow score the highest on it!  [Can you imagine a scientist claiming to have definitive proof a certain race was more beautiful than another?  It's nonsense--beauty is a socially defined concept, and can't be tested through scientific inquiry.] So even if we remove the genetic piece of it, and just ask, "are men or women, given their genetics and how society treats them, more intelligent?"  We can really only answer "more intelligent according to a specific test being administered in a certain way." Lynn claims that because fewer women score in the upper echelon of IQ tests, this means they should be underrepresented at high levels of academia.  But is this the case?  How do we know the skills measured by IQ tests are the same ones that make a great scientist?  So, when we boil it down, the only real scientific question we can answer, given Lynn's data, is "Do men or women score better on a test designed by white men?"  I guess that doesn't make as good of a headline.

Scientists ask questions that can be answered through data.  Ideologues ask questions that fulfill a social agenda.  I have no doubt that there are genetic differences between men and women, between people with European, Asian, and African ancestry, between lefties and righties.  But when we call these genetic differences something socially constructed, like intelligence, and, worse, try to answer them with data that cannot separate genetic and social differences, we remove this inquiry from the realm of science.  Next time Mr. Lynn wants to ask, are adult men or women, raised under current gender norms, better able to answer a specific set of math questions in a specific context, I'll be right there.  Maybe he could join the groundbreaking work of Vesterlund and Niederle, who have shown women who are equally able as men choose to compete less often, and sacrifice their overall payoff as a result.  Until then, let's all stop calling him a scientist.

10 comments:

  1. You could also have mentioned the lovely idea that women are fundamentally less interested in athletic activities than men, and therefore Title IX actually discriminates against men by ensuring equal gender participation rates in high school and college varsity athletics. That's right up there with your pre-1920 politics example.

    ReplyDelete
  2. One of my best friends was a neuroscience major in college, so I asked her about this article when I was talking to her today, since I had just seen it on your blog.

    She said that the guy was right in the sense that the distributions of scores on IQ tests do vary for men and women. The female distribution is more compressed, so it would be accurate to say that at the highest (and lowest) IQ scores, there are more men than women. But she said the means for those two distributions are about the same. So you can't really say, "on average, men are more intelligent than women," since the averages are actually identical.

    ReplyDelete
  3. Hey White_Tree, my problem isn't with the facts of how men and women score on IQ tests, whichever way that comes out. My problem is that he took this to mean "We therefore know there should be fewer women in the sciences, because women are naturally less smart." If we think adult men and women score differently on IQ tests because they are socialized differently, than the problem isn't with women's natural ability, but rather with the education and opportunity they receive. Lynn's way of viewing it means there isn't a problem to be solved with women's under-representation in science (he says it should be the way it is), whereas I think there is.

    ReplyDelete
  4. Characteristics such as sex and race are immutable, and inseperable from their underlying genetic identifiers, meaning there's no way to observe the effect of having a "female" or "black" genotype without having the accompanying socialization that people with those characteristics experience.

    I don't know about sex, but certainly you can do this with race by comparing people with the same race but different genetic makeup, the same race but different circumstances growing up, etc. Not everybody of a given race has exactly the same genotype or life experiences. Look at the studies cited in this editorial:

    http://www.nytimes.com/2007/12/09/opinion/09nisbett.html?_r=1&em&ex=1197435600&en=776bfd8179e46edc&ei=5087%0A

    ReplyDelete
  5. Hey anonymous, the reason I don't agree that this is possible scientifically is that race and gender are visually perceived by people interacting with you. So, even if an African American individual is raised by, let's say, a white family in a white neighborhood, he or she will still be treated as black by people who interact with her or him, and this portion of socialization is inseparable from the genetics that created the visually perceivable characteristic in the first place.

    ReplyDelete
  6. This comment has been removed by a blog administrator.

    ReplyDelete
  7. The social conditions do NOT encourage men to be more intelligent, i would say the exact opposite..
    In smaller ages you are considered "number one", if you a sports-girls orientated person, instead of a "nerd"..The same goes for "popular" and "non popular"(nerdy) girls, but that goes to lesser extend because male nerds are socially outcasted in a much harsher way..
    More women go into the university etc..
    The sexes have areas of intelligence that they show a greater tendency for, so we don't have to make this a gender war..The reason is that by nature we are made complementary to each other..

    ReplyDelete
  8. So your argument is that it is unscientific to perform studies on individuals and draw conclusions about their genotype because the individual is actually in its phenotype form and therefore the only conclusions one can draw are those regarding the phenotype?

    I am not sure that many scientists would agree with you.

    To move on to another one of your points:

    If a study was presented to you that showed that individuals at high levels of academia are more likely to have a high IQ, would you revise your view that the IQ test is "a standard which was designed by the very white males who seem to somehow score the highest on it!" and consider it an accurate measure of Intelligence? (If one's definition of intelligence is such that an individual must must be intelligent to be involved in academia at a high level)

    James

    ReplyDelete
  9. "Anonymous said...

    So your argument is that it is unscientific to perform studies on individuals and draw conclusions about their genotype because the individual is actually in its phenotype form and therefore the only conclusions one can draw are those regarding the phenotype?"

    No, her point is that this experiment lacks controls.

    She is saying is that we would have to "disguise" the phenotype in order to test differences in genotypes. Ex: take "white" people, dye them black and keep them that way till the age of, say, 25, and take "black" people, dye them white, and keep them that way. THEN you can test whether they are smarter or not. This is becoming way too political...

    - The OTHER Anonymous

    ReplyDelete
  10. Are you going to deny that men are better runners, because they are on average taller, and have more muscle mass?

    ReplyDelete

Commenting is now open, but we'd love it if you chose one username so other commenters can get to know you. To do this, select "Name/URL" in the "Comment as" drop down. Put the name you'd like others to see; the URL is optional.

Any profanity, bigotry, or synonyms for "[ ] sucks!" will be deleted. We welcome criticism as long as you're making a point!