July 9, 2010

How did your kid do on the APs?


Scores for the nearly 3 million Advanced Placement test taken by high school students in May are now arriving in the mail. So, in the interests of helping you parents establish your bragging rights, here's the graph of what AP scores equate to in percentile terms. I created last year for a VDARE.com article. It shows how your kid did, but not compared to all the other kids who took the test, who are a self-selected few, but to all the other kids in the country of his or her age (including those who have already dropped out of high school). The brighter the color, the higher the score. This graph starts at the 90th percentile on the left and goes up. An untruncated graph showing the performance of all kids in the country would be ten times as wide.

AP tests are graded 1 to 5 with a 5 supposed to be an equivalent to an A in a typical college's introductory year long course in the subject, a 4 equal to a B, and so forth.

So, if your kid took the English Lit test (the top bar in the graph) and got a 4 (the yellow-orange band), he actually scored at the 98th percentile (or higher) out of all kids his age in the country. If he got a 3 (light gray) in US History the third bar down), he scored in at least the 94th percentile.

Of course, if all students took the test, the number of people scoring 3s, 4s and even 5s would go up. In particular, Red State students don't take APs as much as Blue State students, and whites don't take anywhere near as many APs as Asians.

My 2009 VDARE.com article has lots of graphs on how students do on the AP, overall and by race.

89 comments:

Anonymous said...

I like AP's and SAT II's because it's too rare to have independent assessment of how well teachers are teaching students (don't know how effectively that's changed since NCLB) and how teacher's grades track with students actual performance.

I think grading should be done by third parties, or at least students should be able to opt in for alternate grades or GPA's (for example, proctored standardized exam by a third party).

Hopefully Anonymous

http://www.hopeanon.typepad.com

Underachiever said...

This brings back good memories.

Yan Shen said...

Any statistic documenting the performance of "Asians" is meaningless, given how it lumps together South Asians and East Asians with Southeast Asians and Pacific Islanders. The disparity in performance between the first 2 groups and the last 2 groups is enormous.

SFG said...

You know, I'd like to see you address Red State anti-intellectualism sometime, Steve. It's one of the biggest barriers to our getting more White American engineers and scientists, instead of importing them from China and India.

Anonymous said...

How do you balance red-state anti-intellectualism with the general conservative nature of the engineering profession?

Yan Shen said...

While the graph certainly is informative, the one issue with it is that it gives the impression that certain tests are harder than others, based on the percentiles, even though part of that is inevitably the result of the particular test not being widely taken.

I have no doubt that for the most widely taken tests, the overall percentiles if everyone had taken the test, would not be too different from the percentiles as they're currently presented. But for some of the less widely taken tests, its hard to know how the percentages would shift.

Furthermore, I would wager that a large number of students at elite universities, like in the Ivy League or at comparable schools like Stanford, MIT, Cal-Tech, Duke, etc don't take some of the tests, simply because they know they're not going to get any meaningful credit.

I hate to say it, but most of the elite schools view the AP tests as more or less useless. And the same mentality persists amongst students at these elite schools. The APs are probably the most useful for students at lower rated schools, outside of the cream of the crop, like the Ivies and the Stanfords and MITs.

Anonymous said...

2008 AP Scores by Percentile: All 17-Year-Olds

•Did Not Take •1(F) •2(D) •3(C) •4(B) •5(A)

90% | 92% | 94% | 96% | 98% | 100%


Are these percentages computed before or after about 50% to 60% of all NAMs have dropped out of school?

Anonymous said...

You know, I'd like to see you address Red State anti-intellectualism sometime, Steve. It's one of the biggest barriers to our getting more White American engineers and scientists, instead of importing them from China and India.

I'd say that Blue State nihilistic legalism [to include rent extraction] is a far greater problem - it steers the smart kids into law school, or "finance", instead of something useful, like engineering, or medicine.

[And then, of course, they never make any babies.]

Anonymous said...

Alright, my music theory 5 from 2000 probably makes me pretty boss, but I wonder if that's 'cause no one bothers to take it. Was hoping the macro and micro econ would have been more impressive. The lit and comp scores don't surprise me.

Anonymous said...

I assume your chart is meant in jest, because you can't seriously believe that someone who scores a 1 on an AP exam is in the 90+ percentile.

Also note that most of the elite colleges don't give AP credit, so much of the extreme right part of the bell curve don't bother to take many APs.

Anonymous said...

SFG: But it helps when you need soldiers...

Steve Sailer said...

The only elite college I'm aware of that doesn't give AP credit is Caltech. MIT only gives credit for scores of 5.

Yan Shen said...

At some elite schools you get credit, but not in any meaningful sense. Take for instance the University of Pennsylvania, where I attended. If you got a 5 on either one of the English APs, you get something called an "English Freshman Free" credit unit. Basically, it's not a credit for any particular class, but can be used to substitute for one of humanities requirements. But most students generally end up taking an actual class instead of using their AP credit, to fulfill their humanities slots. Also, if you had gotten a 5 on the Computer Science AB exam for instance, you received credit for CIS 110, a basic intro to Java programming course. But most CS majors skipped that and began with CIS 120, a harder version of that same course. And in fact CIS 110 wasn't even a requirement course for the CIS degree. Furthermore, if you scored say a 5 on the Physics B exam, you got credit for some obscure lab, but engineering students were in fact required to a take a different introductory physics course. So for all intensive purposes, the Physics B AP test would've been useless for say a Penn Engineering student. Other examples abound.

I don't know how it works at other places. But at most elite universities, you basically either get no credit, or you get meaningless credit, for your AP courses. AP credit is a much bigger thing for students at lower ranked universities.

Steve Sailer said...

Right, you don't want to jump directly from high school into second year physics for physics majors at MIT or wherever comparable. You'll get killed.

The point is that these are good national tests of brains and work ethic, so the data is very interesting.

Underachiever said...

"so much of the extreme right part of the bell curve don't bother to take many APs"

Incorrect in my experience. The far right get tons of 5s.

Couchscientist said...

Whether or not elite schools give much to the APs, I'm sure that they take them into account on admissions. A 4.0 means jack at a lot of high schools. 4s and 5s are at least proof of something.

Anonymous said...

Mr. Sailer,
You come across as a bit naive/inexperienced about the achievement environment for hard quant elite performers.

1st, the spread is very wide at the far right, so quite a few incoming freshman would do fine in 2nd year physics for physics majors at MIT.

2nd the vast majority of elite quant performers take a bunch of math/science AP exams and classes. The exceptions are people basically doing the same thing in other guise, like IB or TAMS students (if you don't know what TAMS is you aren't in the quant elite competition cohort, or you're over 40).

Everyone comes out of IMSA, Blair, Thomas Jefferson, etc. with a ton of AP credits, at least they did when I was in college.

The real reason it doesn't matter is because you can place out of any math, science, or language course at any elite university just by arranging a placement exam with the department. It won't give you the graduation credits, or maybe even satisfy a math/science graduation requirement, but anyone at any good school can go straight to the high level stuff pretty easily. Most people choose not to, except those TAMS kids and the occasional oddball in a hurry.

But what do the AP courses really mean for science competition winner types that are doing publication level research and have the background quant literacy for it in high school?

Hopefully Anonymous

http://www.hopeanon.typepad.com

Mitch said...

The far right side doesn't take AP tests? You all are goofy. I teach an AP History review course to East and South Asians, several of whom each year go to a top ten university.

And Steve, you can't use the AP in that way, as vast portions of the middle and high end scores aren't taking the test, and far too many kids at the low end of the bell curve are.

Notice how easy it is to tell which tests are used by schools to pad their "Challenge Index" ratings--to say nothing of getting their kids heavily weighted GPAs despite 8th grade reading and math skills.

The Biology test has more 1s than any other score. The US History test is worse. Macro economics, AP Environmental Science, the list goes on.

In some states (Florida comes to mind) more students get a 1 than any other score.

It's fraud, when you consider the added point the kids get just for having the course on their transcript.

Yan Shen said...

If you wanted to see how the quantitative elites were doing, you'd probably want to focus on something like the American International Mathematics Olympiad team or say the American International Physics Olympiad team.

What's surprising is how the East Asian dominance in the International Mathematics Olympiad(IMO) is greater than the Indian dominance, proportion wise.

Out of the 60 contestant slots for the past decade, from 2000-2009, I counted 26 East Asians, 4 Indians, and 30 whites. East Asians probably make up around 2.5% of the total population, Indians about .8% or so. So even after adjusting for relative proportions, the East Asian representation is at least 2x that of the Indian representation, proportionally speaking. Indians however, have won 8 out of the past 12 spelling bees. :) So, we know that their strengths perhaps lie elsewhere.

I believe that an analysis of the top performers on the Putnam Exam reveals the same East Asian dominance. The IMO and the Putnam are basically the two toughest mathematics contests that any American can participate in, so we definitely know that it selects for the ultra-elite cream of the mathematical crop.

Past IMO medalists and Putnam Fellows have gone on to win the Fields Medal and the Nobel Prize, so it'll be interesting to see how these recent contestants end up doing in terms of real world accomplishments.

Clyde said...

Well. How did yours do?

Anonymous said...

> The APs are probably the most useful for students at lower rated schools, outside of the cream of the crop, like the Ivies and the Stanfords and MITs.

Speaking from first hand experience, this is totally false. All serious engineering types enter Stanford/MIT with oodles of APs. And you get a ton of credits in particular for Physics C.

http://studentaffairs.stanford.edu/registrar/students/ap-charts

Vernunft said...

"You all are goofy."

Whoa, universal generalization does not work that way!

Keep working on the logic.

Bill said...

The point is that these are good national tests of brains and work ethic, so the data is very interesting.

Work ethic?

I took AP tests to avoid classes.

Anonymous said...

Yan Shen,
Interesting research/analysis. You make a good case that East Asians significantly outperform South Asians at elite quant (I wasn't aware that the spread was so dramatic in IMO type competitions).

The prestige of spelling bees in America annoys me as much as it seems to do you.

My personal observations have been that south and east asian quant elites in the USA are about equally high performing (same with WASPs and Ashkis although of course the Ashki performance is amazing given their small global population size) but IMO is more of a global measure, and if your bean counting is right I think it's a significant indication that East Asians are objectively superior to East Indians as elite quant performers.

Hopefully Anonymous

http://www.hopeanon.typepad.com

SFG said...

"SFG: But it helps when you need soldiers..."

Yeah, but the Chinese can just draft them. The same culture that wins in 1950 may not win in 2020, and we have to be aware of that.

As for nihilistic legalism: true, but the blue-state quant elite, like anyone else, wants money, and would do those things anyway. You'd have to take money away from the lawyers and I-bankers (which I'd have no problem with!). I just think there's a lot of untapped potential in farms in Iowa and Montana that winds up fixing cars. Nothing wrong with fixing cars--cars need to be fixed-- but we are losing to China and India. And we are not going to enjoy not being the hegemon in 50 years.

Yan Shen said...

"Speaking from first hand experience, this is totally false. All serious engineering types enter Stanford/MIT with oodles of APs. And you get a ton of credits in particular for Physics C.

http://studentaffairs.stanford.edu/registrar/students/ap-charts"

Clearly you failed to read my post, where I pointed out that even when you received "credit" at an elite university, it was often meaningless. For instance, I pointed out how Penn Engineering students were basically required to take a different introductory course, apart from the obscure lab credit that a 5 on the Physics B exam would've garnered. This was and is the case at UPenn, where I attended. In fact, I'd say that AP courses in the hard sciences probably fail to capture the depth of material that one would learn from actually taking the introductory course at an elite university.

Physics C is probably by far the most difficult of the APs and taken only by a scant proportion of the population, even relative to the other APs. All of the other tests are on a level entirely beneath that of the Physics C exam.

If anything, the AP exams probably underestimate the potential of the far right of the bell curve. Most of the tests are heavily content based, and I knew people who basically slacked off in their studying for the exams, because some of them knew they weren't going to obtain meaningful credit for a large portion of the tests they took. I'd say that the most useful AP test by far is probably Calculus BC, because at virtually every school, you'll get at least a credit for the introductory calculus class with a 5 on the exam at the elite schools, or a 4/5 at lower rated schools.

If you actually bothered looking at your Stanford AP link, you would've seen that they don't even award any kind of credit for large swathes of the AP tests.

Yan Shen said...

@Hopefully Anonymous

The differences are even more stark when you exam the performance of the various nations at the IMO.

India has won 8 golds, 52 silvers, 48 bronzes, and 12 honorable mentions in 21 years of participation.

China has won 107 golds, 26 silvers, 5 bronzes, and 0 honorable mentions in 24 years of participation.

Japan has won 28 golds, 52 silvers, 31 bronzes, and 3 honorable mentions in 20 years of participation.

South Korea has won 38 golds, 54 silvers, 25 bronzes, and 6 honorable mentions in 22 years of participation.

Anyway you slice it up, the East Asian performance at the IMO vastly exceeds that of the Indian performance.

And yes, I definitely think that the spelling bee is overrated. :)
Though apparently some IMO competitors like Evan O'Dorney have done well at both. He won the 2007 spelling bee and won silver medals at the past 2 IMOs while representing team America. He's also on this years 2010 team, which just recently finished its competition, though the official results have yet to be announced.

Yan Shen said...

@Hopefully Anonymous

I should clarify. My original analysis was for the USA IMO team over the past decade. There are 6 contestants per year, so 60 total contestants over a 10 year period. When I examined the 2000-2009 period for the American team, I counted up 26 East Asians, 4 South Asians, and 30 whites. The white category included both gentiles and Jews. My analysis of degree of over-representation was also clearly based off of the proportion of those groups amongst the general American population at large.

The second post shows how the nations themselves do, i.e. China, Japan, South Korea, and India. I think it's fair to say that India lags behind those nations performance wise. Its relatively anemic performance at the elite levels of mathematics competition leads me to wonder whether or not claims about the future rise of India are a bit overstated. Where are the elite upper caste Indians who should supposedly be dominating the toughest math and science competitions in the world?

Anonymous said...

Yan Shen,

I think you're more of a lawyer for east asians than a good faith epistemologist in this space, but I think you're providing useful info.

I like the individual ethnic measure better than the nation measure, because nations vary dramatically in the resources they put into competitions like IMO.

An Indian IMO winner could come out of Singapore or the USA. Of course a Chinese winner could come out of Australia.

Hopefully Anonymous

http://www.hopeanon.typepad.com

Anonymous said...

Off-topic, but this is worth checking out:

http://www.boston.com/bigpicture/2010/07/poverty_within_white_south_afr.html#photo4

http://www.boston.com/bigpicture/2010/07/poverty_within_white_south_afr.html

Anonymous said...

Yan Shen:

The IMO advantage doesn't translate into a real research advantage.

Check this out:

http://www.ams.org/profession/prizes-awards/ams-prizes/morgan-prize

Whites > Indian-Americans > Chinese-Americans (the same pattern repeats in Computer Science contests versus Computer Science Research - check out the ACM Doctoral Dissertation Awards, the MIT Sprowls Awards, the Nevanlinna Prize winners, the Turing Award winners, and contributors to the very elite FOCS conference).

If you look at the Fields Medal, there you have:

Whites >> Chinese-Americans with no Indian-Americans.

With the Abel Prize, it's Whites all the way except for a solitary Indian-American.

I'd wait till the Chinese/Chinese-Americans started producing Fields Medalists in large numbers to actually believe that they hold as significant a "Quant" advantage over Indians as the IMO results indicate.

Anonymous said...

"You know, I'd like to see you address Red State anti-intellectualism sometime, Steve. It's one of the biggest barriers to our getting more White American engineers and scientists, instead of importing them from China and India."

I'd say the lack of tracking in school systems is a bigger drag of the home-grown production of engineers. Also, Blue State kids want to grow up to be political and corporate masters, not guys with green eye shades designing cool stuff like new hydrostatic transmission systems, recoiless auto shotguns, or robotic aircraft. That's the kind of thing Red State males do like to do.

OneSTDV said...

I think AP classes are mostly a waste, other than perhaps Calc.

What's the rush exactly? There's plenty of time to take college courses when you're actually in college.

Anonymous said...

You'd have to take money away from the lawyers and I-bankers (which I'd have no problem with!).

Or round them all up, put them on a ship, sail out into the middle of the ocean [directly over, say, the Marianas Trench], and toss them overboard.

[Cue the old joke about sharks & lawyers & professional courtesy.]

Obama's Mama said...

@Yan Shen
"For all intensive purposes" - WTF? That makes no sense.
The commonly used phrase is "for all intents and purposes."
You obviously did not do well on the English Composition AP!

Curvaceous Carbon-based Life Form said...

" So for all intensive purposes, the Physics B AP test would've been useless for say a Penn Engineering student. "

Yan Shen pops off with a lot of assertions.
But a guy who thinks it's "intensive purposes" rather than the actual "intents and purposes" is hard for me to take seriously.
Does that make me a red-state anti-illectual?
(Out here in the red states, in view of how much the "elite" schools discriminate against smart White non-Jewish boys, we red-state parents can't just assume our boy will get into Penn State, therefore we cannot advise him to just skip the AP courses.)

Anonymous said...

A post like this always generates a comment thread where either an Asian or a Jew derails it into ethnic bragging rights. Do I really give a damn? I got a 4 in English and a 5 in American History back in '76 - are you suitably impressed with my worthiness yet? I passed the foreign service exam on my first try. My oldest took no APs - but his IQ at age 6 on the WISC was 150+ and at age 12 was 130+ - a statistically-insignificant change; his ASFAB score was a 99.

This "my ticket punches beat your ticket punches" crap is soooo tiresome, but hey, the Chinese and Indians are beating the Ashkenazis and all the whites are down the toilet - isn't that Yan Shen's point? So sorry that I'm not all that impressed with your thousands-year-old civilization or your potential ruling of the world. Why do you all get off on this crap? I've always preferred people who can do something useful - like a plumber or an electrician or a soldier - but hey, geeks anonymous, have a good time! In deference to my kid only, this time I'll post anonymously.

texas first! said...

As for nihilistic legalism: true, but the blue-state quant elite, like anyone else, wants money, and would do those things anyway. You'd have to take money away from the lawyers and I-bankers (which I'd have no problem with!). I just think there's a lot of untapped potential in farms in Iowa and Montana that winds up fixing cars. Nothing wrong with fixing cars--cars need to be fixed-- but we are losing to China and India. And we are not going to enjoy not being the hegemon in 50 years.

Leave the people of Iowa and Montana alone!

Thank God Sailer has enough self-awareness not to sound like some of his "quant-elite" blowhard commentators.

Anonymous said...

"I just think there's a lot of untapped potential in farms in Iowa and Montana that winds up fixing cars."

Ehh, speaking as someone in Montana,I wouldn't worry about it. The smart ones know how to use the school system to get out and away. Usually they stay away. Montana is the Land of Duhhh...

Normalize said...

India doesn't do quite as bad when normalized for population, but S.Korea and Japan do very well.

o normalize for population:

Gld Svr Brz Hon Yrs Pop
China 107 26 5 0 24 1,338m
India 8 52 48 12 21 1,183m
Japan 28 52 31 3 20 127m
S.Kor 38 54 25 6 22 50m

#medals/yr/pop * 1000 =

Gld Svr Brz Hon
S.Kor 345 490 227 55
Japan 110 204 122 11
China 33 8 2 0
India 3 21 19 5

Yan Shen said...

Oh okay, I goofed up there. But contrary to your assertion, I scored a 5 on both APs.

Anonymous said...

Got a question for you Steve, why are you such a fringe figure?

Surely if you speak the truth

Yan Shen said...

I meant that I scored a 5 on both English APs. :)

When you type your comments very quickly, mistakes tend to arise. I urge you guys to challenge me on the substantive points raised by my posts, rather than try to nit-pick and point out every grammatical mistake or careless error that I happen to make.

Anonymous said...

"nit-pick and point out every grammatical mistake or careless error"

I don't tend to mistype that badly. "Intensive purposes"? LOL.

You actually thought that was the phrase. That's not a careless error, that's illiteracy.

:wub:

Yan Shen said...

Well considering that white Americans, including Hispanic whites, are 75% of the American population, while Chinese Americans are around 1.2% of the population and Indian Americans 0.8% of the population, it's not necessarily unexpected that white Americans would dwarf Chinese Americans in absolute terms, performance wise, or for that matter Indian Americans.

In fact, the number of Chinese-American fields medalists is 2, while the number of white American fields medalists is 11. And remember, some of the fields medals were won a long time ago in the 1960s and 1970s, before Asian Americans began to do well academically. So white Americans outnumber Chinese Americans 62.5 to 1 population wise, but only outnumber them 5.5 to 1 in terms of Fields Medalists. This isn't of course the most precise or meaningful of numbers, but it's something to work with here.

Yan Shen said...

Actually, the correct count of American fields medalists yields 2 Chinese-American fields medalists, 3 Jewish American fields medalists, and 9 gentile white American fields medalists.

Anonymous said...

Any statistic documenting the performance of "Asians" is meaningless, given how it lumps together South Asians and East Asians with Southeast Asians and Pacific Islanders.


It's not a very fine-grained description. But then, neither is "white", a term used to describe people from over twenty different countries from the Atlantic to the Urals.

Anonymous said...

In fact, the number of Chinese-American fields medalists is 2, while the number of white American fields medalists is 11.



This is the same logical fallacy which underlies discussion of Jewish accomplishment. You're comparing a subset of a population to a population at large. Chinese-Americans are a very select subset of the larger Chinese population, drawn from the brightest among the one billion Chinese. Comparing them to the average white American and acting surprised that they are so intelligent is just nonsensical.

Yan Shen said...

Actually, studies have shown that the average East Asian American IQ is roughly the same as the IQ of East Asians from their respective countries. The typical East Asian IQ is generally quoted as being around 106.

http://isteve.blogspot.com/2009/03/legal-immigrants-hints-of-iq-scores.html

From one of Steve's own posts, we can see that this is roughly the same as the IQ of East Asian American immigrants. So the claim that Chinese Americans must somehow be the cream of the crop of the Chinese people as a whole is hard to sustain.

I have seen plenty of evidence however, indicating that Indian American immigrants are in fact a select subset of the Indian people as a whole. In fact, this very point has been addressed by Steve in quite a few of his posts.

Anonymous said...

Actually, studies have shown that the average East Asian American IQ is roughly the same as the IQ of East Asians from their respective countries.




You're still doing exactly what I called you out on. You're still comparing Asians "from their respective countries" to whites as a homogenized mass. And you still don't seem to understand that you are doing it. So much for your high IQ.

Yan Shen said...

Uh, I have no idea what you're even going on about. Your entire argument was that Chinese Americans were the cream of the crop of the Chinese, and therefore that it wasn't unexpected that they compared favorably to white Americans. My entire post was to show that in fact, the average IQ of the typical East Asian American is the same as that of the typical East Asian from his or her respective country. That would seem to put a significant dent into your cream of the crop argument. Here are you own words, buddy.

"Chinese-Americans are a very select subset of the larger Chinese population, drawn from the brightest among the one billion Chinese. Comparing them to the average white American and acting surprised that they are so intelligent is just nonsensical."

I argued that your claim was patently false. Maybe you'll have some good arguments in the future though.

Anonymous said...

Yeah Whitelandia is quite a country.

Anonymous said...

Out of the 60 contestant slots for the past decade, from 2000-2009, I counted 26 East Asians, 4 Indians, and 30 whites
--

All 4 Indians are brahmins

Anonymous said...

All the Indian math medal winners are brahmins and merchants, 20% of the Indian population
The rest 80% is at blue collar white IQ or below

In the US Indian diaspora, the brahmin and merchants are 50% and the other 50% is at slightly above blue collar white

Anonymous said...

For what it's worth, Professor Flynn researched the children of non-selected, pre-WW2 Chinese immigrants (overwhelmingly Cantonese) and found that they scored approximately the same as white Americans on IQ tests. However, their academic performance was signficantly better. This would seem to indicate that at least southeastern Chinese are roughly the intellectual equivalent of whites, but can somehow succeed beyond what raw intelligence might indicate. Flynn has suggested that stronger motivation, work habits, and family control (among Chinese) might play a role.

Pre-WW2 era Chinese were not that selected, but afterward they were much more so. The percentage of Chinese-American immigrants with college degrees is much larger than the percentage of ethnic Chinese. This indicates that at least in terms of education, Chinese are now pretty selected, though that was not true for older waves of Chinese migrants.

Flynn, Gladwell, and others have suggested that the arduous lifestyle associated with rice farming (long hours, intensity of work, etc.) may have selected for a higher less of industrious among the Chinese, enabling them to outperform relative to IQ. Incidentally, northern China tended to farm wheat, while southern China farmed rice. In the southeast, it seems the productivity of the farmers was the highest and the work year the longest, which enabled southeasterners to establish academies to train their youth on the imperial exams.

This lead to the southeasterns dominating on the exams for a well over a millenia, until they were abolished. There was even a signficant period during which a quota was introduced to keep the southeasterners from being so dominant, which upset the northern elites. The southeasterners were also able to leverage their greater wealth into business, enabling them to dominate most of the domestic and international commerce, even in the north.

It's interesting that both of the Chinese Field Medal winners are from the southeast, as are 6 of the 9 Chinese Nobel prize winners. I think the higher affluence of the region and its exposure to the otuside world play a role, but there's a longstanding view that southeasterners are just plain smart.

Funny thing is that the southeastern Chinese exhibit traits associated with the north Asian Japanese and Koreans, but other north Asian population groups show very different patterns.

Anonymous said...

A more representative sample of Indians would be the Sikhs of the UK or the Indo-Carribeans of Toronto, who arrived as common caste labor. Both groups aren't anything too intellectual, but they do seem to do reasonably well socioeconomically.

India, by the way, is beating China in the software outsourcing and worker visa arena.

Anonymous said...

The Fields has an age-40 cut-off. A large number of Chinese IMO gold-medalists have crossed the age 40 threshold, so the time-lag argument doesn't really hold. The ethnic Chinese (whether in the US/China/where ever) count is far less than that of Whites (with Chinese outnumbering Whites world-wide). Analyzing in such a fashion makes sense given that elite patterns shouldn't be used as a basis to extrapolate to more general populations.

Anyway, I think it's somewhat disingenuous to claim that China's intellectual elite doesn't wind up in the US. You can manually track down where the Chinese IMO/IPhO medalists have done their grad studies if you're interested. Also, do check out the percentage of Chinese on H1Bs who have PhDs - that's *WAY* higher than the proportion of Chinese (in China) with PhDs.

I think the post 1965 Chinese have been the high-IQ ones, while the pre-1965 ones were moderate IQ ones.

About Indians, your point is valid. They're, on the whole, very un-smart, and their US-based is wholly unrepresentative, with the diaspora a couple of standard deviations above Indians in India in IQ.

Anonymous said...

"India, by the way, is beating China in the software outsourcing and worker visa arena."

Doesn't mean a thing. The kind of work outsourced to India has a fairly low cognitive threshold. A high-IQ doesn't lend an advantage when it comes to low-level coding. The real advantages India has vis-a-vis China are:

1. The ability to speak English

2. The ability to bullshit gullible (and short-sighted) American MBAs in English :)

India will see a lot more competition coming out of China within a decade or so.

Anonymous said...

For what it's worth, Professor Flynn researched the children of non-selected, pre-WW2 Chinese immigrants (overwhelmingly Cantonese) and found that they scored approximately the same as white Americans on IQ tests. However, their academic performance was signficantly better. This would seem to indicate that at least southeastern Chinese are roughly the intellectual equivalent of whites, but can somehow succeed beyond what raw intelligence might indicate.



Yes, "somehow".

The fact that academic and real world performance is determined by things besides IQ will forever be a mystery to IQ fetishists.

steve wood said...

The real advantages India has vis-a-vis China are:

1. The ability to speak English

2. The ability to bullshit gullible (and short-sighted) American MBAs in English :)

India will see a lot more competition coming out of China within a decade or so.


Do you think those two advantages are going away any time soon? India has a big lead in both.

English is deeply ingrained at the middle and upper-middle levels of Indian society. At the highest levels, Indians are usually bilingual and furthermore speak with a clear, precise and easily understood accent.

The prevalence of English is a heritage of the Raj, but I think the relatively comprehensible accent might be in part due to the fact that the languages of northern India are Indo-European.

In contrast, I find that even very well-educated Chinese who are near-fluent in English are often hard to understand because of their accents. At the middle level, they can be nearly incomprehensible. (Presumably, this is why there are so many IT tech support lines in India and so few in China. Non-Indo-European-speaking Filipinos seem to have the same problem. I have no trouble understanding Indian tech support staff, but the Filipinos can to tough to follow. It's not the pronunciation of the words themselves, it's the intonation and cadence of speech that's off.)

As for the ability to bullshit whites, well, the lessons of the Raj come in handy there, too.

Anonymous said...

Sikhs in Vancouver and Britain, whose children are slightly below whites in educational and economic attainment, are pretty representative of what regular Indians are capable of doing. This suggests a mean IQ over 90, but less than 100.

There a lot of middle class Sri Lankan Tamil refugees that have ended up in Toronto. They tend to be on part with the middle/upper class white kids in Toronto and have a high rate of univesity attendance.

One more thing about Toronto. Toronto has lots of Indians and other South Asians of varying backgrounds - Tamil refugees, less educated Punjabi Sikh Jats, Indo-Carribean laborers, professionals, East African Indians, etc. They generally tend to be on par, or slightly below, white Torontans in terms of standardized testing.

"Very unsmart" = Wrong. Indians are about as smart as blue collar whites, on average.

I'd say the following castes are at or above the white IQ mean: South Indian and Bengali Brahmins, Marwaris, Khatris, Parsis, Gujarati banias, Chettiars, Sindhi banias. The following castes are in the maybe category: North Indian Brahmins and Kayasthas. The rest are likely below the white IQ mean.

Anonymous said...

A decade or so? In 2020, you'll probably be making the same prediction again.

Indians, by the way, are massively employed in Microsoft, Intel, Cisco, Oracle, and most of the Wall Street firm software development teams. Even moreso than East Asians, who themselves are pretty well represented. These places, especially Goldman and the i-banks, hire on the basis of smarts, work ethic, education, and ability. No low level coding here.

I wouldn't deny that some Indian programmers are more modest in their abilities, but there are many places in the world to outsource to. Why India instead of, let's say, the Phillipines? India = low cost, high IQ workers.

English ability matters, but you can hire bilingual managers to work with a non-English speaking work force. Most of the Indian techies work for Indian managers, who communicate with the Western managers. A software company could easily bring in a team of Mandarin or Faris speaking programmers, and have the bilingual manager communicate with the company.

Anonymous said...

There have been a few studies of
Chinese - Indian comparison at the high end IQ in the USA
--

The first is the Sue and Abe Study in 1984 of UCLA Students based on old SAT

( 2nd generation , regression to mean )

Reported in book, The asian american educational experience by Don Nakanishi

White Verbal = 512
White Math = 577
Indian Verbal = 520
Indian Math = 606
Chinese Verbal = 473
Chinese Math = 612

---

The second is a survey in Silicon valley, Only high end,
Chinese = 50% MS, 35% PhD, 1% MBA
Whereas Indians = 66% MS, 12% PhD, 20% MBA

--

The third is a study to see what percentage of US high school students get MD seats

( 2nd generation, regression to mean )

http://www.unmc.edu/Community/ruralmeded/ethnicity_gender_and_rural_pract.htm

1 out of 19 Indian ( 50% Upper caste, 50% Kulak Peasant ) High School students get into MD
1 out of 41 Chinese High School students get into MD
1 out of 200 White High School Students get into MD school

--

Per La Griffe du Lion, verbal IQ has a much bigger impact on real world

Anonymous said...

H-1B workers make about 50-60K per year in salary. For a decent paying and "low cognitive threshold" job, American companies could hire lots of people here. Such as you.

Anonymous said...

The Japanese-Americans are a completely non-selected community. If we want to assess Asian elite quant ability, let's look at their representation. Anybody have any figures?

Chinese in Hawaii are mostly pre-WW2 in origin. We could look at them too.

Yan Shen said...

I would be willing to bet that the overall average verbal SAT score of East Asians is similar to that of the average white verbal SAT score, even though the East Asian SAT verbal scores are depressed by the immigrant effect.

Remember, finding data for a single school for a single year is fairly meaningless. You need comprehensive statistics. I've found scores at specific schools for specific years where the average Korean SAT verbal score exceeded that of the white SAT verbal score. But I wouldn't go from that single data point to arguing therefore that the average East Asian SAT verbal score was higher than that of the white one.

http://professionals.collegeboard.com/profdownload/SAT-Percentile-Ranks-by-Gender-Ethnicity-2009.pdf

Now we know that the SD for the Asian distribution is extraordinarily high because the Asian category lumps together under-performing Pacific Islanders and Southeast Asians, along with East and South Asians. This is why on the Verbal portion of the SAT for 2009 there were more Asians than whites at the highest levels of performance, but also more Asians than whites at the lowest levels of performance. It's not unreasonable to conclude that removing Southeast Asians and Pacific Islanders would result in a higher average Verbal SAT score relative to whites. I think that even if you then remove South Asians, the verbal score for East Asians alone would be fairly similar to that of whites. And this would be the case even with slightly depressed verbal scores due to the immigrant effect.

Anonymous said...

Indians, by the way, are massively employed in Microsoft, Intel, Cisco, Oracle, and most of the Wall Street firm software development teams. Even moreso than East Asians, who themselves are pretty well represented. These places, especially Goldman and the i-banks, hire on the basis of smarts, work ethic, education, and ability.






Ha ha ha.

No, corporate America hires on the basis of "diversity" first and foremost. That's why they love Indians - they're the good blacks.

And once Indians get in the door and have a say in hiring, no more whites will be hired.

Anonymous said...

Chinese in Hawaii surpassed white Americans in income as far back as 1950. Chinese in the mainland took significantly longer. Some of this is due to the greater level of discrimination on the mainland, but partly it might be due to the ethnic mix. Chinese-Hawaiians came from coastal districts of Canton with extremely rigorous traditions of rice farming and academics/business, whereas many (not all) of the pre-WW2 mainlanders came from further inland where the pace of farming and level of achievement was much more modest. Maybe not too surprisingly, inlanders and coast dwellers showed highly different levels of achievement even in the U.S. This was true of their children here and even true in Latin America, the Phillipenes, and the Carribean.

However, due to the dynamics of immigration and marriage, the inlanders weren't able to bring over the wives much. So they were supplanted eventually.

I think a cultural hypothesis makes sense, but I wouldn't discount the possibility of ecological selection.

Anonymous said...

It's not unreasonable to conclude that removing Southeast Asians and Pacific Islanders would result in a higher average Verbal SAT score relative to whites. I think that even if you then remove South Asians, the verbal score for East Asians alone would be fairly similar to that of whites





So if we look at a subset of Asians, we (may) see that they perform as well as the homogenized white average?

The mystery is why you think this sort of cherry-picking of data is in an way meaningful.

Anonymous said...

Data from Toronto school districts

Indians are lumped with muslim Pakistanis and Bangladeshis
and reflect a more average mix
and not many brahmins

Good grades in 6th grade

East Asian = Reading - 78%, Writing - 83%, Math - 86%

White = Reading - 76%, Writing - 75%, Math - 72%

South Asian = Reading - 68%, Writing - 74%, Math - 68%

South East Asian = Reading - 69%, Writing - 77%, Math - 72%

Middle Eastern = Reading - 55%, Writing - 57%, Math - 54%

Latin American = Reading - 51%, Writing - 55%, Math - 42%

Black = Reading - 47%, Writing - 55%, Math - 37%

Curvaceous Carbon-based Life Form said...

"As far as I can tell, whites have roughly the same IQ more or less. In other words, there isn't a split between different groups of whites IQ wise in the same way that there's a huge difference between the IQs of Southeast Asians as opposed to Northeast Asians."

Well, then, you haven't done your homework.
Ethnic Germans have avg. 1Q of 107.
Ashkenazis (some argue whether classifying them as White is correct, but they certainly are counted as White when Blacks complain about White overrepresentation in good jobs/schools) 110

Meanwhile, Irish have average IQs of 93.

http://www.timesonline.co.uk/tol/news/uk/article697134.ece
http://en.wikipedia.org/wiki/IQ_and_the_Wealth_of_Nations

107-93 = 14 points. Essentially the same gap between Germans and Irish as between average 100 IQ White Americans and average 85 IQ Blacks.

(You find the same pattern in America. Massachusetts at the top, West Virgina at the bottom, nearly a full standard deviation in avg. IQ.)

So for you to compare East Asians against the average for all Americans is no more fair to us than if we compared Germans to all Asians in order to dismiss your claims of braininess.

Yan Shen said...

@Curvaceous

You might want to try a better list of numbers.

http://en.wikipedia.org/wiki/IQ_and_Global_Inequality#National_IQ_and_QHC_values

These numbers are the result of many IQ tests done repeatedly throughout the years, and not merely one particular test during some particular year.

There's no way that the IQ of ethnic Germans is 107. I remember an article in Times which claimed that the average Japanese IQ was 111, based off of one test.

http://www.time.com/time/magazine/article/0,9171,1101861006-143333,00.html

"Indeed, a report by British Psychologist Richard Lynn published in May 1982 indicated that over the past generation Japan's mean national IQ score has risen 7 points to an average of 111, well above the American norm of 100."

So we know that IQ scores based off of a single test should be taken with a grain of salt. The IQ of Germany was listed at 102 for the 2002 study and 99 for the 2006 study for the link I posted above.

Also, see here.
http://en.wikipedia.org/wiki/Race_and_IQ#History

"According to Richard Lynn and others, racial differences in IQ scores are observed around the world.[43][44] A commonly-cited review by Richard Lynn lists IQ scores for East Asians (105), Europeans (99), Inuit (91), Southeast Asians and Amerindians (87 each), Pacific Islanders (85), South Asians/North Africans (84), Non-Bushmen sub-Saharan Africans (67), Australian Aborigines (62) and Bushmen (54).[45][46][44][47][43]"

On average Europeans are listed at about 100. There is however a significant gap between East and Southeast Asians.

Yan Shen said...

In other words Curvaceous, your fundamental thesis is false. There aren't significant differences in IQ between the Germans as compared to say the British or the French, once you perform the relevant meta-analysis and examine numerous IQ tests conducted throughout the years. In fact, the 2006 numbers from IQ and Global Inequality do just that, and what it reveals is that virtually all of the European nations cluster around the 98-99 to 101 range. In other words, most of the major European ethnicities are similar to one another IQ wise and average around 100 or so. This is remarkably consistent with the general psychometric literature.

What's different about the Asian category is that it lumps together racially distinct groups such as Northeast Asians, Southeast Asians, and Pacific Islanders. Unlike the different European nations, these groups do in fact differ from one another significantly IQ-wise.

Anonymous said...

In fact, the 2006 numbers from IQ and Global Inequality do just that, and what it reveals is that virtually all of the European nations cluster around the 98-99 to 101 range.




That's not true. According to Lynn, Croatia has an average IQ of 90.

Anonymous said...

From Lynn's data on the IQ of nations:

Austria 102
Germany 102
Italy 102
Netherlands 102
Sweden 101
Switzerland 101
Belgium 100
UK 100
Hungary 99
Poland 99
Spain 99
Denmark 98
France 98
Norway 98
Russia 96
Portugal 95
Slovenia 95
Romania 94
Bulgaria 93
Ireland 93
Greece 93
Croatia 90


By way of contrast, these are the figures for East Asian countries.

South Korea 106
Japan 105
Taiwan 104
China 100
Malaysia 92
Thailand 91

There's a twelve point spread between the lowest and highest European countries. There's a fifteen point spread between the lowest and highest East Asian countries.


Yan Chen is incorrect in asserting that "virtually all of the European nations cluster around the 98-99 to 101 range".

Anonymous said...

On average Europeans are listed at about 100. There is however a significant gap between East and Southeast Asians.


It's really uncanny the way you continue to make the same mistake over and over again, even after it has been pointed out to you.

But perhaps the problem is simply that you don't know the names of the countries in Europe.

Anonymous said...

What is funny, is that although the Indians in the US are subselected as cream of the cream of the crop, the average American thinks they are all geniuses.

I think their elite will begin to do better in math competitions, they are becoming more visible in these academic competitions in the last few years. They won't reach the asian levels ever however, IMO.

Anonymous said...

The NE Asian Japanese and Koreans are racially very different from the southeastern Chinese, who cluster with other SE Asian populations. Perhaps there was a similar mechanism selecting for high IQ in both places, but there's no "East Asian" population group that's completely homogenous in IQ.

Mongols, Uzebks, Kyrgz, Turkmen, Tajiks, and Kazakhstan(of North Asian descent) don't display elevated levels of cognitive ability, but Koreans and Japanese do. Those of the Malay and other SE Asian subgroups are about a SD below the white mean, but the Cantonese and Fujianese are at or slightly above the white mean. Even overseas, people of different regions of China display markedly different outcomes.

Really, East Asia is a very diverse place. Depending on where we are talking about, the mean could be anywhere from the low 80s to slightly over 100.

In the case of Asian-Americans, many are Japanese or Korean. Others come from Pakistani or Indian subgroups, who tend to be highly educated and high performing. Many of the Chinese immigrants are highly educated, with the less educated being mainly Fujianese or maybe Cantonese. Lots of SE Asians (Viets, Thais, Malays, Filipinos) are actually ethnic Chinese or of partial Chinese ancestry. It might be fair to say that about 80%of Asian-Americans are from the high performing groups, with the other coming from different backgrounds.

OhioStater said...

Parents in New York like diversity except in their schools:

http://www.nytimes.com/2010/07/13/nyregion/13appraisal.html?hp

Yan Shen said...

Uh Malaysians and Thai are Southeast Asians smart guy.

Try the most recent 2006 data from IQ and Global Inequality. China comes in at 105, along with Japan, Korea comes in at 106.

Most of the major European nations come in at around 98 to 101.

Anonymous said...

In terms of IQ, it seems like there's an East Asian intensive agriculturalist state pattern that exists in states that have taken up many Chinese institutions and written language for some time and that only these people have elavated IQs, whether elevated relative to European or elevated relative to the world average, while other East Asians might do well compared to much of the rest of the world, but still aren't particularly impressive. I find this quite fascinating.

Anonymous said...

"Uh Malaysians and Thai are Southeast Asians smart guy."

Um, yeah. That's why I grouped SE Asians with Malays.

The IQ tests administed in China were done on urban populations from Beijing and Shanghai. Both cities have historically been very well educated, and affluent, by the standards of the country. Even during the era of Communist rule, it seems Beijing had something like 3x the income of the rural areas. Shanghai, which tests somewhat higher than Beijing, was also built by southeasterners.

The Chinese government has, for well over a millenia, had a policy of excluding peasants from living in the large cities, due to the the threat of political instability. Cities tended to be preserves for the wealthy, connected, and elite. As poverty stricken as the country could be, the major cities were always quite orderly and prosperous. Even after the Communists took over, they continued this policy. They also expelled many younger people (in the millions) from the urban areas to the countryside (during the Cultural Revolution, I believe), eventually allowing back ony those that could prove useful to the urban party apparatus. The Communists have even had policies of slum demolition and giving urban residence permits to those from educated, skilled, or entrpranuerial backgrounds (ie rural masses no need apply).

Not at all representative of the general population of migrant workers and slum dwellers.

It also is worth noting that historically, the overwhelming bulk of successful imperial exam candidates came from the southeast or, in fewer cases, the urban north. It's not surprising the urban northerners did okay on the test, as the cities were home to a disproportionate share of gentry and bueracrats. For the southeasterners, who were mostly from peasant backgrounds, it's extremely impressive how well they did. Especially in contrast to the northern and central Chinese peasantry, who were poorly and sometimes not at all represented in the ranks of elite exam takers. The southeasterners also dominated much of the domestic trade, including even in the north, and international trade.

I'm not saying the northerners and peasants have low IQ, but China is in no way homogenous enough to generalize without extensive testing of all subpopulations. There's also the issue of the poor performance of north Asian and Central Asian border states, and the extreme rural-urban income inequality. To lump in diverse China with homogenous Korea and Japan is a mistake.

Anonymous said...

You jokers might want to look at the F_st statistics for Asians. The genetic distance between two randomly selected Chinese, or between a Chinese and a Korean or Japanese, is less than the F_st between Northern and Southern Europeans. This suggests that the IQ spread will not be as large as in Europe. (Note this is not a statement about the standard deviation; it's a statement about the population averages in different regions.) It doesn't prove that this is the case, but it's more meaningful than the stupid hand waving so far on this thread.

BTW, there have been some comments circulating to the effect that the Beijing government got the "g gospel" some time ago. Yes, you can find Raven's scores for all Chinese cities and provinces if you look hard enough. I will let you geniuses figure out what the results say.

Curvaceous Carbon-based Life Form said...

Hey, Steve,

If your 9th-grade boy gets a 5 on the AB/BC calc AP test, is that good?

Being's all us Red-Staters are ignorant cowpokes, I'm not plugged into The System enough to know.

Anonymous said...

Japanese are similar to Mongolians. So I think we can also expect Mongolia to become a first world industrial power sometime very soon, right?

Anonymous said...

http://www.unmc.edu/Community/ruralmeded/ethnicity_gender_and_rural_pract.htm

1 out of 19 Indian ( 50% Upper caste, 50% Kulak Peasant ) High School students get into MD
1 out of 41 Chinese High School students get into MD
1 out of 200 White High School Students get into MD school

Because this has large scale data and uses the entire Med school population and entire high school population, it is not cherry picking data and does not rely on fatness of extreme right tail as for awards. All these awards are in the +4 SD range

From this data, 2nd generation, regression to mean,

Mean white IQ = 100, SD = 15
0.5% of whites get into med school
and for bell curve, this Med school entry bar is 38IQ above white mean

Med School Entry bar for whites, Indians and chinese = 138

For Chinese Americans, The top 2.4% get into Med School, This is 29 points above the chinese mmean, which makes Chinese American IQ = 138 - 29 = 109

For Indian Americans, the top 5.2% get into Med school, this is 24 points above the Indian average

American Indian Average IQ = 138 - 24 = 114

Anonymous said...

@Yens comment about elite upper caste, and various peoples comments about 81IQ Indians

The current mens chess champion for the past decade is Vish Anand , South Indian Brahmin
The current womens chess #2 is Humpy Koneru, South Indian Merchant

The first Asiatic Science Nobel was CV.Raman, South India brahmin, working in a lousy colonial era college
Subrahmanya Chandrasekhar, South Indian Brahmin, also working in lousy colonial college, discovered Type-1A Supernova Standard Candle, for which he got a Nobel later

No 81 IQ population or even 90IQ generic population wins Science Nobels or World Chess Championships
There are also a billion blacks, Latinos and Muslims who dont achieve this

rob said...

Med School Entry bar for whites, Indians and chinese = 138

That's about 15-25 points too high. The average physician doesn't have a 138 IQ. Plus, most people who get into med school didn't score 800 on the math section of the SAT, which would expected for such a high bar.

Anonymous said...

Hey Mr. 138, many many gifted students don't go to med school. Most actually won't. I think a med school typically attracts students in the top 5-10%, intelligence-wise, plus a signifant number of less qualified students but with an excellent memory.