Stephen Pinker revisits the Lawrence Summers controversy, and I’m reminded of a good point a friend made recently: There’s apparently ample literature that whatever intrinsic differences in math (or whetever else) ability exist between men and women, they’re dwarfed by the differences that appear when people believe there’s some kind of biological disparity. In other words, give two randomly selected groups of women the same math test, but tell one group beforehand that research shows that women typically do less well on this sort of test. That group will, apparently, proceed to do less well.
So I wonder: Assume for a moment there really is some cognitive difference such that there are more men who are capable of the highest-level performance in math, say. Might it be the sort of thing that, even if neuroscientists and other academics acknowledge among themselves, they should agree not to talk about too often or loudly in public?
10 responses so far ↓
1 Will Wilkinson // Feb 4, 2005 at 3:08 pm
How about if you tell them (1) that the average woman scores higher on math tests (true, I think); (2) that the a higher percentage men do dismally and fantastically well (also true, I think); or (3), both — that women on average do better, and are less likely to perform at sub- and sperhuman levels?
People are discouraged when you confront them with all sorts of truths. Just how silent should we be about the truth?
2 Grant Gould // Feb 4, 2005 at 4:34 pm
Heck, for that matter there are lot of truths about test performance, and I doubt that the effect in question is limited to gender. Tell them that rednecks, drunks, and lobotomy victims statistically score badly. Gender’s not special, and there’s no reason to focus your morally obligated truth-telling on it.
I’m sure that we can to our hearts content construct moral edge-cases in which morally obligate behaviour appears morally prohibited and vice versa — indeed, I’d bet a stack of dodgy railway switches and suspiciously undersized life-rafts on it.
The fact is, we’re under no particular obligation to manufacture such situations for ourselves. If there is a central truth of hypotheticals, it is that we seldom find ourselves in them. Real life presents myriad options for us to shape and fiddle the truth to our own ends, and avoid stark choices between stupid options.
–G
3 Chuck // Feb 4, 2005 at 4:43 pm
This “intimidation by truth” effect is known in education jargon as “stereotype threat.”
This is Claude Steele’s big idea. ( http://www.personal.psu.edu/users/t/r/trc139/ ) I think it is fairly well accepted among people who study the black-white test gap, although I’m not sure if they agree on the effect size.
4 Tim // Feb 5, 2005 at 8:19 pm
Summers made his comment at a private retreat of Harvard academics. IIRC, there were no recordings of his remarks precisely because the event was set up to allow frank discussion of controversial issues. If it wasn’t for the grandstanding of certain overly excitable female professors, we never would have heard about the controversy at all.
It wasn’t the Larry Summers and Steven Pinkers of the world who were shouting about this from the rooftops, and they’ll hardly be to blame if more girls flunk their algebra midterms this Spring.
5 Tim // Feb 5, 2005 at 8:20 pm
Could you turn off the “require email” feature please? I’m just going to give you a garbage address.
6 Katherine // Feb 7, 2005 at 9:43 am
It’s good to know I’m getting through to you on occasion 🙂
7 joe o // Feb 7, 2005 at 3:53 pm
The fact that a math paper with a women’s name was ranked one point out five lower in quality than the same paper with a man’s name amazes me. How could you fight through that prejudice.
8 Lauren // Feb 7, 2005 at 11:26 pm
We shouldn’t silence the truth, even when it’s nasty, but if we’re aware that any gap (or lack thereof) between races/genders gets larger because of this perception, maybe people can let go of the idea that standardized test scores are nothing more than objective measures of merit. In other words, if women underperform on tests because they think they’re destined to perform poorly, perhaps we can get past the idea that those women are “underqualified” for schools/jobs based on those scores.
9 Katherine // Feb 8, 2005 at 11:39 am
Lauren — Are you suggesting that we disregard test scores for both genders, or just that we give /women’s/ scores less weight when making decisions about qualifications?
10 Lauren // Feb 8, 2005 at 3:45 pm
I don’t know how we deal with test scores. One of my main responses to people who oppose affirmative action is always that test scores underreport the “qualifications” of women & people of color. How do we correct for that? I don’t know. But covering our eyes and claiming that test scores are all we’ve got and are all we should use seems like the worst possible solution.