verbal iq

Not to beat a dead horse, but I think both effects are at work.

I’m not sure that we can interpret what Verbal(SD) (less than) Quantitative(SD) means without knowing more about the underlying populations. In absence of any other info this would be more convincing, but Occam’s razor says the bigger effect is due to the large percentage of non-native English speakers in CS&Eng being significantly handicapped on the GRE Verbal test no matter what their underlying verbal IQ distribution looks like. If many CS/Eng grad schools don’t consider GRE verbal scores, this would further disconnect reported GRE Verbal scores from underlying Verbal IQ.

Markku,

Several comments in a previous post and others have addressed the oft-repeated claim that Asians have a lower verbal IQ. Someone here cited a chart by La Griffe that was actually demonstrated the counterpoint.

“The data supporting a lower East Asian verbal IQ is pretty weak, if not nonexistent. If you look at SAT verbal scores by ethnic group, the Asian average is not much below the non-hispanic white average, even though a sizeable number of Asian test-takers don't come from English speaking homes (many are even recent immigrants). If you correct for that environmental handicap, the verbal average may well be higher for Asians. (It's also true that the Asian category that ETS uses includes some south-east asian groups like Hmong or Filipino that drag down the NE Asian averages -- it's been remarked before that the variance of the collective "Asian" group is extremely large.)”

I suspect that the “Asians have high visio-spatial but low verbal IQ” meme may be more a result of anecdote, perception and cultural bias. Forget the high percentage of immigrant Asians, even native-English American-born Asians I know who ace verbal IQ-like exams would pale next to an Al Sharpton or Mohammed Ali in a typical public exchange in the minds of the masses. Was it the movie “American Pimp” where a black guy said he never saw no white pimp because they lack the charisma? Same thing at work but even stronger.

Aston,

ETS has made the GRE quantitative section far too easy. It’s easy enough so non-math/science/engineers can feel some mastery it but far too easy to provide good discrimination among the math/science/engineers. As a result, it probably has less predictive power among those where it is a given that everyone clusters around the very top.

As someone else mentioned here, the GRE Quantitative section is little more difficult than the SAT Math section. Trying to use the GRE Quant section to predict performance in a PhD physics program is like giving a 4th grade math test to H.S. seniors and expecting the results to predict college GPA.