<< an account moreover which makes no attempt to state
with what level of certainty that hypothesis can be put forward). >>
The econometric analysis in the "Bell Curve" is so poor that that he'd never pass an econometrics course, let alone get it publishedn in a second tier econ journal.
Factor analysis is just another statistical technique to identify factors that (may) have some correlation. I don't believe that you get any measures of association for the individual components other than a simple correlation coefficient, and as Rakesh notes, no confidence intervals.
But the more important point is that for all their vaunted statistics, in fact, they are rather "primitive." For example, why did they even use factor analysis? I suspect that there was so much multicollinearity in their "explanatory" variables that they couldn't get a robust estimator from the lot - this is Econometrics 101! Not to mention that his kind of cross sectional data is likely replete with heteroscedasticity, structural weakness, and plain ole "specification error" (I love this one because there really is no way to "disprove" that you don't have such an error -despite Wald and other tests). Specification error is fatal because your parameters can never converge to their "true" values - that is actually the most obvious, fatal problem with Murray, and indeed ANY use of econometrics. You know, Keynes had a devastating critique along the lines of the early econometric model builders. Keynes was one smart SOB with a sharp eye for statistical bullshit
Suppose Murray wrote a book ONLY about comparing the IQ of Japanese and Jews, which found that, ON AVERAGE, the former had a "statistically significant" higher mean score than the latter. He then went on from there and did his usual narrative about lower earnings, etc. Now suppose Murray were (by some miracle) revealed to be BLACK!? What kind of reaction would the mainstream press be?