The Bell Curve
I finished reading The Bell Curve last week. Like Bowling Alone, it's packed with data, though the presentation isn't quite as engaging. There are only so many regressions comparing the influence of IQ and socioeconomic status that I can go through before they lose their allure. Still, it was worth it to have so much data in one place. I'll just reproduce one of the many things I learned: a list of the validities (that is, correlation coefficients) of various predictors of job performance ratings.
The source for this data can be found here. (Thank you, Google Scholar.)
In the last four of twenty-two chapters, the authors Herrnstein and Murray did write a little bit about what they thought the policy implications of their book should be. However, the rest of the book is straight-up science. I found their science very careful and that they were humble in drawing conclusions.
With my overall praise made clear, I will venture my one scientific quibble. In their chapter on the demography of intelligence, they talk about "dysgenic pressure" of "0.8 [IQ] points per generation" in America (modulo uncertainty about the Flynn effect), presenting plenty of evidence that, on average, less intelligent women bear more children and do so at younger ages. E.g., in 1992, "the overall average IQ of American mothers was a little less than 98." They also cite some studies from the 1980s and earlier that looked at age cohorts of both men and women. However, in trying to assess the dysgenic pressure in America in the early 1990s (the book was published in 1994), they only consider female fertility patterns and immigration. They completely ignore whether or by how much this dysgenic pressure is mitigated by the likely positive correlation between intelligence and fertility among males. Perhaps there was no data for them to talk about, but in that case I think they should have mentioned their ignorance of this effect. Likewise, if there is a reason why they could discount male fertility patterns, then I wish they would have explained it.
| Predictor | Validity |
|---|---|
| Cognitive test score | .53 |
| Biographical data | .37 |
| Reference checks | .26 |
| Education | .22 |
| Interview | .14 |
| College grades | .11 |
| Interest | .10 |
| Age | -.01 |
In the last four of twenty-two chapters, the authors Herrnstein and Murray did write a little bit about what they thought the policy implications of their book should be. However, the rest of the book is straight-up science. I found their science very careful and that they were humble in drawing conclusions.
With my overall praise made clear, I will venture my one scientific quibble. In their chapter on the demography of intelligence, they talk about "dysgenic pressure" of "0.8 [IQ] points per generation" in America (modulo uncertainty about the Flynn effect), presenting plenty of evidence that, on average, less intelligent women bear more children and do so at younger ages. E.g., in 1992, "the overall average IQ of American mothers was a little less than 98." They also cite some studies from the 1980s and earlier that looked at age cohorts of both men and women. However, in trying to assess the dysgenic pressure in America in the early 1990s (the book was published in 1994), they only consider female fertility patterns and immigration. They completely ignore whether or by how much this dysgenic pressure is mitigated by the likely positive correlation between intelligence and fertility among males. Perhaps there was no data for them to talk about, but in that case I think they should have mentioned their ignorance of this effect. Likewise, if there is a reason why they could discount male fertility patterns, then I wish they would have explained it.
posted by sedenion on 4/19/2005
0 Comments:
Post a Comment
<< Home