about me
art
biz
Chess
corrections
economics
EconoSchool
Finance
friends
fun
game theory
games
geo
mathstat
misc
NatScience
... more
Profil
Logout
Subscribe Weblog

 
In the May 2006 American Economic Review (Papers and Proceedings from AEA annual conference) there's an article about attrition in Economics PhD programs (see outline here). What caught my eye was they had the average GRE scores for the top 48 phd programs, and the average GRE Quantitative score was a 775, and average verbal of 568. Considering most of the students were not US citizens, much of the verbal score was probably biased downward due to lack of familiarity with English. Nontheless, I found a GRE to IQ translator here, and this suggests--conservatively because they equally weight the Quantitative and Verbal scores--an average IQ of around 139, or the 99.4 percentile.

There wasn't a big difference between the top 15 schools that dominate research and the rest (4 IQ points or so), though there wasn't much data on schools ranked lowed than 48th place. Kind of scary, when you think about all the dumb stuff that goes on in these places. But I think it explains why economics research about the underclass is so lousy: they have no clue how a person with an 85 IQ thinks or behaves, and are too egalitarian to assume they think differently.

As my MBA friends would say, too bad we can't learn common sense (or how to dance, etc.)
tc (guest) meinte am 29. Jun, 00:53:
There is an extreme amount of "test prep" on the Q these days. Applicants will practice the math section until they are fairly certain of getting an 800 (and most do). I'd say that the verbal section is actually a more accurate measure of true IQ, since there's much less preparation for it (and it's harder to game in any case). 
A (guest) meinte am 29. Jun, 02:00:
Dream on
As purportedly clever and evidence based folk, I'd encourage you to think carefully about any attempt to convert GRE's to IQ scores. As a research psychologist, I know of no study that has attempted, much less succeed, in assembling valid evidence to support such a translation of GRE scores.

And perhaps the difficulty economists have with understanding the economic realities of your supposed intellectual inferiors in the "underclasses" reflects more about their own cognitive and other deficits.

But by all means, please, feel free to flatter yourself publicly all you like. It is quite entertaining and telling for the rest of us. 
HedgeFundGuy antwortete am 29. Jun, 03:02:
Well, I'm no research psychologist, but it passed my smell test, and I do have the link. I know Mensa used to allow SATs as proof for admittance, so I presume there's been a mapping, and it seems reasonable that SAT and especially the GRE are pretty "g" loaded. In any case, if the average IQ is 100, and the average US college graduate IQ has an IQ of 115 (1 stdev above average), it isn't hard to think that a US phd program is 1 stdev above that, especially in quantitative fields, in better phd programs. 
dsquared (guest) meinte am 29. Jun, 13:45:
test practice bias
But surely anyone who has taken the GRE in order to be admitted to a graduate program, would have practiced GRE tests before taking it? I know from taking the GMAT test that practice makes a huge amount of difference to your score. 
Paul N (guest) meinte am 30. Jun, 05:02:
I actually would expect the GRE analytical section, which I guess they got rid of in 2002, to be the best correlate with IQ. It's sort of like how they got rid of analogies on the verbal SAT to make it less of an IQ test and more of a how-much-did-you-study test (and more like the ACT). I actually kind of enjoyed the analytical section in a sick way because I always got them all right although it kind of hurt my brain after doing it for an hour in a row.

Discussions about IQ make me depressed because I always do well on tests and IQ stuff but I haven't been very successful at anything I've done since college. Being in the top 0.1% or better doesn't make you cool or effective, it just makes you a huge nerd. 
scientologist (guest) antwortete am 1. Sep, 10:50:
the gre is an iq test
no research is required to demonstrate the /sat/etc is or is not an iq test for anyone who knows what an iq test is.

first, there are many tests calling themselves iq tests made by professional test makers. they don't agree, but they do correlate. the ravens-stanford-binet correlation is .7. the gre correlates at least as well.

if a test correlates with an iq test at .7 or above it is an iq test, because iq tests do not correlate with each other any better

second, iq tests arose when binet discovered that many tests have high correlations, ie measure the same thing. put them together in a battery and you get an iq test.