In addition in the previously linked slides he seems to present the regression to the mean of IQ scores in children as a fact, but in this blog post it's obviously conjecture.
There will be regression to the mean in IQ as long the correlation between the IQ of children and the average IQ of their parents is less than one. A lot of research has found this to be the case. Hsu is not breaking new ground here. Galton discussed regression to the mean in the 1800s.
Really? Galton discussed regression to the mean of IQ scores in children of highly gifted parents that long ago, or did he just discuss height? Does Hsu have some source for his assertions of hard facts, for example the .6 factor etc., or are they just based on fiddlings with numbers in turn based on his assumptions of how it might work, with an admittedly simplified model, while demonstrating lack of knowledge of how the calculations ought to be made, and while allowing that the real world might not be that way? Stating some basic ideas about regression to the mean in general is not enough to convince me in this context.
(Also, is there some independent corroboration, besides from the publisher of the test, about how g-loaded the current version of the SAT is?)
I do detect a bit of an agenda on the part of Hsu. It's not really based on disagreement with anything he believes; I at least would agree that if racial differences in intelligence exist, and are valuable to discuss, we should do so without fear.
