December 21, 2010

PISA school test scores by ethnicity

My big graph showing how the four main American races would do compared to 64 other countries in this year's PISA reading test scores is up at VDARE.com.

74 comments:

Jason Malloy said...

Steve,

Here are the PISA 09 US ethnic data for science and math.

I tried to send them to you, but your email wasn't working.


Best,
Jason

Anonymous said...

Fujian, Zhejiang, southern Jiangsu, Chaozhu, and coastal Guandong were the provinces that produced a disproportionate share of China's top academic scholars. This southeastern strip of land had very fertile land that enabled farmers to produce a high standard of living, providing them with excess capital to fund the building of a large number of academic institutions. It also provided merchant clans with the assets neccessary to monopolize much of the domestic and international trade.

At one point, this small, though densely populated, region of China was producing almost all the top exam takers in China. The authorities in Beijing, resentful of this and desiring a more equitable regional distribution of competitive candidates, imposed a quota on the southeast. Some southeasterners, especially from Zhejiang, tried to evade the quota by relocating north.

Shanghai was predominately populated by people from Zhejiang and Guangdong. It also drew a significant proportion of China's gentry. Northerners relocated to the city as well to work as laborers, but it's not clear how many of them brought their families or permanently settled down.

Today's Shanghai is comprised of much the same population, but many well educated Chinese and affluent entreprenuers have been able to get residence permits to live in the city. Much of the slum population has been ejected by the government's aggressive urban renewal.

It might be fair to say that Taiwan is fair reflection of the intellectual capabilities of the southeastern Chinese, as it was founded by middle class peasants from Fujian province, who arrived with small amounts of capital and began to farm the land. Shanghai, I believe, is probably slightly inflated because of policies that result in the dull moving out and the bright moving in.

Anonymous said...

"Much of the slum population has been ejected by the government's aggressive urban renewal."

There are 21 million people in the municipality. It's hard to flush so many people of average intelligence out.

Anonymous said...

The Chinese that initially came to the U.S. and Cuba originated from different regions of Guangdong. Those that came from the coast, which had a very strong history of high performance on the imperial examinations and a vibrant mercantile tradition, did quite well in the U.S. and Cuba. They formed families and supplied much of the pre-WWII Chinese-American population. They predominately spoke the Son Yee and Zhang Shan dialects.

The Chinese that came from the poorer inland sections of Guandong, which had no history of participating in the imperial exams and had an economic base dependent on fishing, were generally much poorer than their coastal brethren. They spoke the See Yee dialect. Since the U.S., at the same, had a specific immigration provision that prevented Chinese immigrants from bringing over their wives unless they were "merchants" (ie they had to own a business with a substanial asset value), the See Yee formed very few families. They did form the large majority of the Chinese population for a while. Which is partly why the Chinese were previously stereotyped as being poor, bachelor laborers during the early 1900s.

In Hawaii, whose Chinese were mainly from the coast of Guandong, the Chinese made the move from laborer to middle class businessman very quickly. Very different from the mainland, where the See Yee bachelors dragged down the community average until most of them died off.

Anonymous said...

Not every average person has been pushed out, but the slums have been razed and many of their residents have been pushed out. Shanghai's per capita income is 3 times the national average for a reason.

Many of the 21 million residents of Shanghai are temporary workers too. Not families.

My guess would be that Taiwan is pretty representative of the intellectual potential of the region, more than Shanghai.

SFG said...

I'm surprised USA-White is so high. Are we really doing better than the Germans and French? I just don't buy it.

nsam said...

Nice job.

American students have a slight advantage in being more familiar with testing (of the sort PISA is), esp since NSLB. So one could argue that the American scores are slightly inflated, more so for the lower performing sub-groups (where the thrust of SOL testing has focused on).

TH said...

The data explorer Jason linked to was pointed out to you in the comments earlier, Steve. You should read your commenters!

Here are the American math and science results by race:

Mathematics:

White 515
Black 423
Hispanic 453
Asian 524

Science:

White 532
Black 435
Hispanic 464
Asian 536

-----

I'm surprised USA-White is so high. Are we really doing better than the Germans and French? I just don't buy it.

German and French figures include Arab, Turkish, etc. immigrants, so they aren't comparable.

sabril said...

"I'm surprised USA-White is so high. Are we really doing better than the Germans and French?"

To make it an apples-to-apples comparison, you would need to compare USA-White with German-White and France-White.

Both of these countries have a lot of Muslim immigrants from Turkey and North Africa. The percentage is probably even higher among 15-year-olds.

josh said...

SFG,

I don't think we have data on German and French whites.

Formerly.JP98 said...

Completely fascinating. Thanks for doing this.

I haven't read your whole analysis yet, but I was very surprised to see that Asian Americans beat white Americans in reading but not science or math. I would have expected the reverse.

Jason Malloy said...

Using the NCES data explorer tool to make charts, here's race/ethnic data for:

PISA reading 2000 and 2009.

PISA science 2006 and 2009.

PISA math 2006 and 2009.

TIMSS math 2007 (grade 4).

TIMSS science 2007 (grade 4).

TIMSS math 2007 (grade 8).

TIMSS science 2007 (grade 8).

PIRLS reading 2006.

robbie1687 said...

Fascinating article, Steve. One of your best imho.

Anonymous said...

Mostly the chart you put together adds up to a picture of incoherence in the PISA results.

For example, the score of USA Blacks is essentially the equivalent to that of Serbia, Romania, and the Russian Federation. Given that the latter countries are pretty much Caucasian through and through, how does that make any sense, if the PISA tests are really measuring with any accuracy "g"? Apart from some serious special pleading, attempting to attribute to certain groups of Caucasians vastly different underlying distributions on "g", how do those numbers get explained away?

I think that these sorts of results reinforce the general difficulty in making fair cross cultural comparisons of cognitive ability.

wmhde said...

Steve:
do home schoolers take this test? My sense is that most homeschoolers are over the 80 percentile? what about private schools, like Philips Andover?
I can easily imagine one of them blowing it off for SAT practice.....

Black Death said...

I too was surprised by the strong performance of the USA-White population, which outscored every other Caucasian nation except Finland. With respect to the influence of ethnically non-native populations in Europe, specifically Germany, the CIA World Fact Book describes the German population as follows:

Ethnic groups:

German 91.5%, Turkish 2.4%, other 6.1% (made up largely of Greek, Italian, Polish, Russian, Serbo-Croatian, Spanish)

....

So it wouldn't seem to make much difference, especially since many of the Eastern Europeans in Germany are of German ancestry.

Felix said...

I asked my Chinese girlfriend about her education in China. She told me that starting from middle school through high school, she was at school from 7:30AM to 9PM, 6 days a week, with an occasional Sunday thrown in once in a while. Factor in the Asian obsession with education compared to white diffidence, and you start to get an idea that if IQ plays a part in these results, it's likely a very small part.

Anonymous said...

Note that black students in the United States scored higher than students in several mainly white countries (Romania, Bulgaria, Montenegro, Albania, Argentina, Uruguay, and perhaps Azerbaijan).

Peter

Felix said...

Oh yeah, one other thing she told me. In China, there is an intense selection process that starts at the middle school and then high school levels. That is, kids are tested and then the top 10% or so are sent to "the best" middle schools and then again they are tested at the high school level and the top are sent to the best high schools. In other words, students in China are segregated even at an early age, and some schools are composed of only elite students.

Since I doubt every single high school in Shanghai was tested this leads to the question: WHICH high schools in Shanghai were tested for this PISA study? Clearly, it would skew results if the Shanghai authorities steered the testers into predominantly elite high schools.

Felix said...

I asked my Chinese girlfriend about her education in China. She told me that starting from middle school through high school, she was at school from 7:30AM to 9PM, 6 days a week, with an occasional Sunday thrown in once in a while. Factor in the Asian obsession with education compared to white diffidence, and you start to get an idea that if IQ plays a part in these results, it's likely a very small part.

Anonymous said...

Could the USA breakdown by race be used, by holding those scores constant, to project our likely overall score decline from 2010 to 2050, assuming Census Bureau projections about changing American demography turn out to be correct?

Mitch said...

Steve, great job. I've been looking for exactly that data for a while. Thanks!

Anonymous said...

"I haven't read your whole analysis yet, but I was very surprised to see that Asian Americans beat white Americans in reading but not science or math. I would have expected the reverse."

The Asian category lumps together East Asians along with Southeast Asians and Pacific Islanders. East Asians have roughly a 1 SD IQ advantage over Southeast Asians/Pacific Islanders. Lumping those groups together would be as meaningful as lumping blacks and whites together into one racial category. Despite the fact that Sailer realizes the huge shortcomings of the aggregate Asian American statistics, he routinely trots them out anyway.

Anonymous said...

But then I’m not an educator.

Wrong.

Albertosaurus

Truth said...

So why exactly does "USA-Black" outscore Bulgaria, Uruguay, Romainia, Montenegro, and Argentina again?

Have we come up with a rationale...er...um...reason for that one yet?

USA-B is only one point behind Chile also. I guess Chile, Argentina, and Uruguay are temporarilly back to being inhabited by "Mess-kins" on this board again, after being Southern Scandinavia for so long?

Anonymous said...

Are we really doing better than the Germans and French?

In fairness to the Krauts & the Frogs & the bloody Limeys, you really ought to do what Steve did for the USA - remove the 40% to 50% consisting of Muslims and then re-calculate.

Eurabia über alles!

Chief Seattle said...

Very interesting, Steve. Now I know why you wrote the blog on graphs.

For something that doesn't exist, IQ is pretty darn stable. A Bell Curve reader in 1995 could have placed bets on the 2009 outcome and come pretty close. The main surprise is the rapid increase of wealth in coastal China, but even that hasn't altered the underlying patterns.

Anonymous said...

“a lot of Muslim immigrants from Turkey and North Africa.”

When I looked a few years ago, American Whites scored EXACTLY the same as native born Swedes.

I honestly don’t know why you are surprised. American whites have far better schools. Schools in Europe have less resources and are even more infected by PC and “modern” education than American schools.

American whites have a strong protestant work ethic (even the Catholics) and the same genetic raw material as European whites.

American whites also produce about 50% more in yearly output than French and German whites. Native born American whites get almost half the world’s scientific Nobel prizes. I could go on.

American whites get a bum rap because they are intentionally lumped in with other Americans in crime, education, and other social statistics. This way the left gets to be both anti-American and anti-white. As a single group you do spectacularly.

Anonymous said...

"Black Education Disaster
by Walter E. Williams"

Over at the Libertarian site the blame on low education scores is "the poor education received by blacks".
http://www.lewrockwell.com/williams-w/w-williams63.1.html

Hopefully the writer of that article will read your blog and expand his thinking.

Anonymous said...

"To make it an apples-to-apples comparison, you would need to compare USA-White with German-White and France-White."
Indeed, and also for Dutch-White, Swedish-White, etc. The Vdare post indicates that New Zealand more readily provides this ethnic-breakdown info, so maybe the other countries have it as well. Just a question of how to get ahold of it.

Anonymous said...

Asians trounce whites at reading? That's odd. But of course the government, in its infinite wisdom, considers Hispanics to be white, so that must lower all the white scores a notch.

Anonymous said...

To make it an apples-to-apples comparison, you would need to compare USA-White with German-White and France-White.

You also need to keep in mind that "white" in America does not even mean white.

Severn said...

Mathematics:

White 515
Black 423
Hispanic 453
Asian 524

Science:

White 532
Black 435
Hispanic 464
Asian 536



What jumps put at me here is that we basically have two groups rather than four. Hispanics are much closer to blacks than to whites or Asians. And the white-Asian difference is negligible.

In fact Asians appear similar to Jews in that their presence in colleges and universities exceeds that what their raw academic performance would indicate.

It would be nice if the HBD-sphere would get off its obsessive fixation with IQ and academic performance and explore some of the other characteristics which affect different groups.

OneSTDV said...

So our education system is working great. What a surprise.

Anonymous said...

Asians outscored whites by 9 points on the math, in the United States. However, in the international results, the Asian nations outscored the Western nations by 45 points. So, evidently, there is something to the idea that the rigor of the East Asian academic system is conducive to high scoring.

Or, to put it another way, cram school is not a waste of money....

I think maybe the AA average is dragged down a little by Pacific Islanders and SE Asians (their proportion of the AA population is not that large), but it's also boosted up by having a lot of highly educated Chinese and Indian engineers and scientists in this country.

Anonymous said...

Would Finnish students be advantaged due to linguistic reasons? Obviously, for English speakers, Finnish is a difficult language, but is Finnish easier for Finnish speakers than English is for English speakers? I have heard, for instance, that spelling rules in Finnish are much more consistent. In English, you can regularly come across a word you don't know the meaning for, but is this prevalent in Finnish?

Anonymous said...

You know, Trinidad and Tobago is a middle income nation that's around half Indian and has a decent score. If we assume a mean score in the mid 300s for the blacks (which I got from looking at middle income South Africa's scores on the TIMSS), then Indians there might be performing slightly below the Euro mean. Like mid 90s. Not great, but not too shabby either.

I think the tremendous variance among both the Euro and Asian countries, and even the good performance of US NAMS, suggests that schooling system has a fairly strong impact on standardized test scores. That runs contrary to the determinist, IQ-only argument.

Though, the gaps within the U.S. are still large. So maybe IQ is a key factory, but environment can close at least some of the gap.

If we convert PISA scores to IQ, US Blacks scored 12.6 IQ points below US whites. Hispanics scores 8.9 points below. Traditionally, blacks have scored about 16.5 points below whites (ie 1.1 Standard deviations) and Hispanics, who are more heterogenous group, varied more.... in general Hispanics were about 12-13 points below US whites. What's interesting is that both Hispanics seem to have closed the gap by roughly the same amount (3-4 IQ points). So I guess all our efforts at closing gaps do work to some extent. Approximately, 1/4 of a standard deviation.

Anonymous said...

Some Chinese commentator said the academic pressure in China is super intense, even by East Asian standards. He attributed Shanghai's high score, in comparison to the other East Asian superstars, to the extreme educational culture of the country.

Interesting that Japan seems to be slightly behind the other East Asian countries. Also, South Korea is ahead of Taiwan by 44 points on reading, but 3 points on math. Sort of surprising....

Anonymous said...

Formerly JP88, you'll note that the science and math data Steve referenced were from 2003 and 2006. The same data is available from 2009 following Jason Malloy's link, though TH has already pulled the data.

What is interesting is that there seems to be a data anomaly in the results from different years of the PISA.

Has anyone noticed that from Steve's graph, all nonwhite minorities in the US exhibit the same general trend. That is the 2006 science scores are slightly lower than the 2003 math scores. The 2009 reading scores are then a major improvement. However for US whites, this trend is not evident. Instead the 2006 science scores are moderately higher than the 2003 math scores. The 2009 reading scores are then a marginal increase over the 2006 science scores.

I'm at a loss as to what is causing this anomalous trend in the data. Perhaps sampling biases introduced or removed in the different years of the PISA test. The math scores are roughly analogous from 2003 to 2009 for all ethnicities. Blacks and Hispanics have increased their science scores a modest amount compared to Whites this year. The real surprise is that Asians now score higher than Whites across the board in reading, math, and science while this was not true for 2003 and 2006.

Deckin said...

Great stuff, as usual. But here's the surprising thing: Daniel Koretz, of the Harvard Ed School (the belly of the PC beast if ever there was one) came to exactly this conclusion a few years ago in his book Measuring Up. There, if I recall correctly, he noted that if Hispanic and African American students were subtracted out, US scores in science would put them second in the world and math, around fifth.

Where you really surpass the good professor is in demonstrating that our African American and Hispanic students outperform their world averages. We should be singing the praises of our educational institutions (along with screaming about our immigration policies), not beating up on the par saps naive enough to volunteer for combat duty in urban schools.

The most interesting question is why does someone like Koretz get away with this while you get harassed by the SPLC?

Jeff said...

I don't buy the white - Asian data. Remember, our Asian data includes more than just East Asians. The white gap to American East Asians must be pretty high across the board. The average white student is fairly stupid given that most white students are products of the left half of the white bell curve. Anyone looking for a point of white pride is going to have to look elsewhere. Quite simply, white people are screwed: they are soon to be surrounded by fairly dull immigrants and white culture is well down the road to tattoos, pit bulls and juvenile dispositions.

I teach at ground zero for our nation's future. My school is 1/3 white, 1/3 black and 1/3 Hispanic. In this environment, nothing productive can happen because the groups are not equal in potential and behavior. Thus, the only product is the status quo. No one can push the students because it will result in too many black and Hispanic failures (plenty of whites will fail too, but no one cares about white kids). No one can enforce discipline because it will result in too many black referrals. Thus, even if your worst student is a white kid, you are limited in your options because all options have been rescinded due to the overwhelming, unbelievable, and insane hooliganism of about 15% of the black student population.

Now, if my school above is representative of the future, what is going to happen in "whiter" areas? Quite simply, those areas are going to reduce standards because their performance will be so high relative to ours. When such a natural performance disparity exists, the natural result is that the winner begins to reduce effort; there is no need to be perfect when your competition gets an award for hand washing.

Elbrac said...

http://www.americanthinker.com/2010/12/the_college_diploma_fraud.html

Diplomas turned into derivates, just like on Wall Street.

Anonymous said...

"Asians outscored whites by 9 points on the math, in the United States. However, in the international results, the Asian nations outscored the Western nations by 45 points. So, evidently, there is something to the idea that the rigor of the East Asian academic system is conducive to high scoring. "

Are you an idiot or what? Asian Americans include East Asians and Southeast Asians/Pacific Islanders. The difference in IQ between these two groups is almost 1 SD. Please try to think a bit more carefully before you spout off with your idiocy on this blog.

Anonymous said...

@Severn

Asian Americans include both East Asians and Southeast Asians/Pacific Islanders. Lumping these two groups together would be like lumping together blacks and whites. The IQ differential between East Asians and Southeast Asians/Pacific Islanders is roughly 1 SD. Steve Sailer realizes this, but he repeatedly and dishonestly trots out the aggregate Asian American scores, even though he knows that most of his readers will assume that the Asian American numbers refer to East Asians.

@Anon

Uh no, Southeast Asians make up quite a large chunk of the US Asian population. Check the wikipedia. There are more Filipino Americans than there are Chinese Americans. There are more Chinese Americans than Japanese/Korean Americans combined. There are also as many Vietnamese Americans as there are either Korean or Japanese Americans.

Look at the distribution for Asian Americans for a test like the SAT. Because of the lumping together of various groups, the SD in the Asian American scores is extremely high, which is why you see a vast over-representation of Asian American scores at the high end.

Anonymous said...

Are you an idiot or what? Asian Americans include East Asians and Southeast Asians/Pacific Islanders. The difference in IQ between these two groups is almost 1 SD. Please try to think a bit more carefully before you spout off with your idiocy on this blog.


I don't know if you are illiterate or just plain retarded. Probably both. Reading your post is making us dumber by the minute. Soon enough, we'll reach your level of intellectual capacity.

Let me requote what I said:

"I think maybe the AA average is dragged down a little by Pacific Islanders and SE Asians (their proportion of the AA population is not that large), but it's also boosted up by having a lot of highly educated Chinese and Indian engineers and scientists in this country.
"

Pacific Islanders and other low performing SE Asians are only about 8 percent of the AA population. Filipinos form a large part of the population, but most of them are middle class and they have a good college attendance rate, with many of them being of ethnic Chinese descent. Vietnamese-Americans are only slightly below the higher performing AAs in test scores and tend to do well academically.

This is all somewhat counterbalanced by the huge number of AAs that are really well educated and whose children likely boost AA PISA scores. For example, the percent of Chinese-American and Indian-American people with a degree far exceeds their native populations, which would bring up the AA average.

I'm not sure if you're smart enough to understand all this, but please try before you make a fool of yourself.... again......

Anonymous said...

Vietnamese-Americans achieve pretty good test scores. Comparable to whites, in math at least. Filipinos do decently too, especially since a lot of them are Chinese-Filipinos and even the remainder are really well educated. Filipino-Americans have a really low poverty rate. Lots of other southeast Asians in this country are ethnic Chinese too.

That's 1 SD bullcrap is only true of PIs, Cambodians, and Lao. These groups are less than 1/10th of the Asian population in America.

In Australia, SE Asians are actually considered a high performing group. Why? Because almost all of them are ethnic Chinese.

Steve Sailer realizes this, but he repeatedly and dishonestly trots out the aggregate Asian American scores, even though he knows that most of his readers will assume that the Asian American numbers refer to East Asians.



Steve is more honest than you. If you look at the IQ test results for new immigrant children, they show East Asians at 106 IQ. Southeast Asians at 104 IQ. Indians at 111 IQ. These impressively high IQ results are partly a result of all the super educated technical workers we get from Asia.

Anonymous said...

A better question is why East Asian parents send their kids to all these jukus and crammers if these places are ineffective. Do Asian people just like to waste money and time?

"I asked my Chinese girlfriend about her education in China. She told me that starting from middle school through high school, she was at school from 7:30AM to 9PM, 6 days a week, with an occasional Sunday thrown in once in a while. Factor in the Asian obsession with education compared to white diffidence, and you start to get an idea that if IQ plays a part in these results, it's likely a very small part.
"


"In fact Asians appear similar to Jews in that their presence in colleges and universities exceeds that what their raw academic performance would indicate.

"


Nisbett found the Asians score about the same as white Americans on IQ tests, but vastly outperform them on achievement tests. So, for whatever reason, Asians are doing better than raw cognitive ability would suggest.

Richard Flynn noted the same trend. AAs (Japanese and Chinese) descended from pre-WWII immigrants scored slightly lower than white Americans, but still did a lot better on most academic and professional metrics.

Anonymous said...

This study compared Greeks with Chinese, from 8 to 14 years of age, on measures of processing efficiency, working memory, and reasoning. All processes were addressed through three domains of relations: verbal/propositional, quantitative, and visuo/spatial. Structural equations modelling and rating scale analysis showed that the architecture and developmental patterning of the various processes are basically the same in the two ethic groups. The Chinese clearly outperformed the Greeks in all tasks addressing visuo/spatial processing, from processing efficiency through working memory and reasoning, but neither in g nor in processes where the two groups have equivalent experience. This advantage was associated to the massive practice in visuo/spatial processing that is required to learn the Chinese logographic writing system. The implications for general theory of intelligence and intellectual development are discussed.

Anonymous said...

@Anon

You my friend, are obviously unfamiliar with the psychometric literature. According to Richard Lynn...

"According to his calculations, the East Asian cluster (Chinese, Japanese and Koreans) has the highest mean IQ at 105, followed by Europeans (100), Inuit-Eskimos (91), South East Asians (87), Native American Indians (87), Pacific Islanders (85), South Asians & North Africans (84), sub-Saharan Africans (67), Australian Aborigines (62), and Kalahari Bushmen & Congo Pygmies (54).[44][45]"

http://en.wikipedia.org/wiki/Race_and_IQ


Southeast Asians and Pacific Islanders score significantly lower than East Asians on IQ tests. They also make up a fairly large percentage of the Asian American population. The ethnic Chinese who immigrate from Southeast Asia are obviously going to score higher on IQ exams, because they're racially East Asian. Now, if you'd actually like to engage with my honestly here, I'd be more than willing to clarify your utter ignorance of the facts.

Anonymous said...

A better question is why East Asian parents send their kids to all these jukus and crammers if these places are ineffective. Do Asian people just like to waste money and time?

Maybe the kids need a glorified babysitting service?

TH said...

For example, the score of USA Blacks is essentially the equivalent to that of Serbia, Romania, and the Russian Federation. Given that the latter countries are pretty much Caucasian through and through, how does that make any sense, if the PISA tests are really measuring with any accuracy "g"? Apart from some serious special pleading, attempting to attribute to certain groups of Caucasians vastly different underlying distributions on "g", how do those numbers get explained away?

It seems probable that Southeast Europeans have lower IQs than other Europeans. Moreover, Serbia and Romania have sizable Roma minorities that aren't too smart.

Russia, on the other hand, is a large collection of different ethnicities. It is far from all-white.

I don't think anyone has claimed that the PISA tests are measuring just g. The quality of the school system obviously counts. It might be informative to look at not just the mean scores but also the SDs of different groups and the proportions of students at the highest levels of PISA achievement.

It would be nice if the HBD-sphere would get off its obsessive fixation with IQ and academic performance and explore some of the other characteristics which affect different groups.

I would surmise that those characteristics are related to the different distributions of personality types across races. The problem is that compared to IQ tests, personality tests are not very reliable, and in intercultural comparisons they are pretty much useless. HBDers concentrate on IQ, because we just don't know much about racial differences in other behavioral characteristics than IQ.

With respect to the influence of ethnically non-native populations in Europe, specifically Germany, the CIA World Fact Book describes the German population as follows:

Ethnic groups:

German 91.5%, Turkish 2.4%, other 6.1% (made up largely of Greek, Italian, Polish, Russian, Serbo-Croatian, Spanish
)

Those are for the whole population, whereas the PISA tests concerned 15-year-olds.

Furthermore, from Wikipedia:

Over 99% of those of age 15 and above are estimated to be able to read and write. However a growing number of inhabitants are functionally illiterate. The young are much more likely to be functionally illiterate than the old. According to a study done by the "University of Bremen" in coorporation with the "Bundesverband Alphabetisierung e.V.", ten percent of youngsters living in Germany are illiterate and one quarter was able to understand only basic level texts.[15] Illiteracy rates of youngsters vary by ethnic group and parents socioeconomic class.

none of the above said...

Anonymous:

In societies where there are a limited number of slots for admissions to top schools, and admission to a top school pretty much guarantees your place at the top of the social order, I think there's a pretty classic rent-seeking situation going on. Each student (or his parents) will spend almost the whole value of the benefit of getting into that top school, in order to get there.

So you can get millions of kids doing unreasonable hours of cramming for the high-stakes test that determines their future, even if studying for that test does them no lasting good at all. The test could be on conversational Klingon, and those kids would still spend hours every day cramming for it, because for each individual student, it would make sense to max out their score so they got into a top school.

I suspect that's going on in various Asian countries with the cram-school culture, and to a lesser extent, among American kids shooting for Ivy League admissions. Though I also wonder if there's some historical cultural link between high-stakes tests for tracking students, and the old Chinese mandarin exams.

none of the above said...

It's also possible that Chinese kids do better in school than their IQ would suggest, not because of cultural differences, but because of genetic differences in personality. I suppose the only way to really untangle this would be with some kind of trans-racial adoption study.

Tom V said...

Anonymous (Yan Shen):

Are you an idiot or what? Asian Americans include East Asians and Southeast Asians/Pacific Islanders. The difference in IQ between these two groups is almost 1 SD.

There he goes again, our Mongolo-Nordicist Yan Shen. Let's see. Thailand placed 5th at the 2010 Math Olympiad, ahead of Japan (7th). Vietnam is 11th, ahead of Taiwan at 19th.

Are the Thai kids mostly of Chinese descent? Maybe.* The Chinese diaspora are certainly well represented on the Indonesian (36th) and Filippino (74th) teams.

Now how many of these kids do you think will end up studying in the US at some point? And staying? And how many of their peers (less talented, but similarly of the high-IQ minority background)? Quite a lot, I'd say. Many of those Southeast-Asian Americans that you so despised are actually your fellow Chinese. Of course, they probably wouldn't admit it to you.

Interestingly, the Asian-American label also encompasses South Asians, who are even less related to Northeast Asians than Southeast Asians. Why don't you complain about including them? It's not just because you like their high scores, is it?

* The Thai kids' last names are not as conspicuously Chinese as I had expected. (All Chinese Thais have by now adopted Thai last names, but those names are often distinctive. Not the case here, except for the adult team leader) In fact, some of those names seem very ethnic-Thai. And those overexposed photos are useless for our purpose.

Tom V said...

After all these years, Truth is still playing with his imaginary friend who insists that nature is everything and nurture is nothing.

Kind of cute, actually.

Anonymous said...

Researchers tested nearly 6,000 students majoring in science and engineering at seven universities -- four in the United States and three in China. Chinese students greatly outperformed American students on factual knowledge of physics -- averaging 90 percent on one test, versus the American students’ 50 percent, for example.

But in a test of science reasoning, both groups averaged around 75 percent -- not a very high score, especially for students hoping to major in science or engineering.



That suggests that Chinese are roughly the equivalent of Americans in raw ability, even if the Chinese know more and solve problems better.

Extrapolating from this study, that might be why Taiwan can do so much better than Greece on standardized testing, but both groups seem to have similarly advanced economies.

Anonymous said...

@Tommy V

Look at any of the SAT score reports for the past few years. The SD for the Asian American scores is usually somewhere around 20% larger than the SDs for the other groups. Asian Americans are over-represented at both the high and low end of the bell curve. This is clearly the result of East Asians/upper caste Indians being lumped together with Southeast Asians/Pacific Islanders.


http://professionals.collegeboard.com/profdownload/sat-percentile-ranks-by-gender-ethnicity-2010.pdf

La Griffe du Lion notes this point accordingly as well.

http://www.lagriffedulion.f2s.com/women_and_minorities_in_science.htm

"An attempt was made to fit the "Asian" MSAT distribution for 1987 and 1992 (the years for which we had access to the ethnic distribution data) to a sum of Gaussians. Since the elite Chinese, Japanese, Koreans and South Asians made up about half this group, the Gaussians were given equal weight. (See Figure 3.) The least-squares fit suggests that this Asian quartet performed with a mean of approximately 627 for 1987 and 641 for 1992. The remaining Asians had mean scores of approximately 426 (1987) and 444 (1992), not unlike American Indians. "

Now, if you or any of Steve Sailer's WN readers would like to dispute this point with me, I'd be more than glad to engage with you.

Anonymous said...

"Since I doubt every single high school in Shanghai was tested this leads to the question: WHICH high schools in Shanghai were tested for this PISA study? "

Sampling flaws are probably not severe if present at all. 5000 students in Shanghai were tested in sample designed to be represenative. The test is lavishly funded and Australian education bureaucrats had oversight.

Anonymous said...

"Since I doubt every single high school in Shanghai was tested this leads to the question: WHICH high schools in Shanghai were tested for this PISA study? "

Sampling flaws are probably not severe if present at all. 5000 students in Shanghai were tested in sample designed to be represenative. The test is lavishly funded and Australian education bureaucrats had oversight.

Anonymous said...

I think maybe the AA average is dragged down a little by Pacific Islanders and SE Asians (their proportion of the AA population is not that large)

Guys, it seems like this depends on how Steve aggregated his PISA data. I'm assuming he wrote NE Asian there for a reason...

Severn said...

I don't buy the white - Asian data. Remember, our Asian data includes more than just East Asians.


Our "white" data includes more than just whites, it includes Hispanics. And if we're going to be as specific as you are being with "East Asians", then the whites we want to look at are the north-west and central European whites.

Severn said...

Nisbett found the Asians score about the same as white Americans on IQ tests, but vastly outperform them on achievement tests. So, for whatever reason, Asians are doing better than raw cognitive ability would suggest.



Obviously. But the question of "why" would seem to be of pressing importance.

It's also worth noting that Asians don't seem to achieve all that much in the great achievement test that is life. All those billion-plus high IQ East Asians have remarkably little to show for their high IQ.

Anonymous said...

The high US black scores relative to certain "white" countries is interesting.

For certain countries (Argentina, Russia) I have no insights.

Regarding the Balkan countries, here is a bit of historical info which may or may not be relevant: during the centuries in which the Balkan countries were under Ottoman Turkish occupation/oppression, the native Balkan peoples were subjected to a recurrent demographic culling, in which the most promising boys were abducted from their families and transformed into the janissary corps, who did not form families.

This culling may have some bearing on present-day Balkan IQ levels; or perhaps a better historian will tell us it doesn't. Also the effect may be offset by comparable selection effects among the US black population, I wouldn't know.

If anyone knows more about these complex effects please chime in.

I just thought I'd put that data point out there for others to consider.

Polymath said...

On my blog
here
there is a discussion of the results.

In that discussion, RebelliousVanilla blames the poor Romanian results on a terrible educational system and points out that at the high end Romanians are well-represented:

"Considering that here schools just teach you binge drinking, I’m not surprised. The whole education system here should be abolished since it provides no benefits.

Basically, everyone here knows what they do in spite of the system, which is amusing because it does a huge triage among the haves and have nots in terms of intelligence. This is why if you’re average, you’ll turn out dumb, if you’re smart, you’ll turn out well. This is why we get to do well in international contests, but horribly in terms of averages....Also, sadly, I suppose a lot of it has to do with the communists killing off intelligent people too because they were not equal."

JSM said...

How come Chinese toasters stop working in a few months?

How come battery-operated toys made in China stop working before Christmas day is even over?

How come "made in China" is synonymous with "crap"?

And, note, this is not a singular anomaly.

Back in the day, back when "made in America" was synonymous with "quality," before all American manufacturing got offshored, the synonym for "crap" was "made in Taiwan."

Are the Han Chinese dull-witted or do they just not give a crap?

Felix said...

Sampling flaws are probably not severe if present at all. 5000 students in Shanghai were tested in sample designed to be represenative. The test is lavishly funded and Australian education bureaucrats had oversight.

It's not simply a matter of sampling flaws. Like I said previously, from what I've been told by someone who's actually been through the Chinese educational system, schools in China are organized in tiers according to student abilities. The smarter students are sent to specific schools, and the less bright are segregated to other schools.

This idiosyncracy of the Chinese education system leads me to qestion just how accurate the Shanghai results are. Did the Australian test administrators even take into account the fact that all Chinese schools are not equal? If they did not, then the Chinese authorities could have easily gamed the testing by holding it in predominantly elite schools, which would have GREATLY impacted the results.

Assuming the Australians took the stratified nature of Chinese schools into account, how would they then apportion the 5000 students between them so that the sample of students tested was truly representative of the whole student population?

Anonymous said...

La Griffe gave the gaussians equal weight, which I think is a dubious assumption. It presumes that Vietnamese and Filipinos perform significantly under the white American average. If you look at test scores and other educational or economic stats, this isn't true.

I'd believe that Filipinos and Vietnamese would drag down the Asian average, but not by a lot. Especially since a lot of them are ethnic Chinese.

It's also relevant that the Asian average is boosted by a lot of AAs being from really well educated families.

Here's a better comparison.... Shanghai is an overwhelmingly middle class and upper class city, at least in terms of the resident population with kids in the school. It's also populated by high performing East Asians.

Now take the Asian-white gap on the math PISA and assume that about half the AA population (Filipinos, Vietnamese, PIs) pull down the AA average a lot. Assume about 1/2 a SD below the white average, as La griffe presumes. The other half of the AA population is high performing.

If we make these assumptions, as you do, then the high peforming AAs have a math mean of 580. Really good, but still 20 points below Shanghai. That's the equivalent of 3 IQ points, which is not trivial.

3 IQ points represents the Asian-white IQ gap often given, by people like Charles Murray and Lynn.

Anonymous said...

Asians are academic high achievers, beyond what IQ would indicate, because they are good at getting their work done. They also are more obediant to their parents and teachers. Diligence and obediance are two qualities that can help a lot in the class room, which then translates into a good job and stable life.

Behaviorally, Asians are completely the opposite of blacks or Pacific Islanders or Middle Easterners. To borrow a Whiskey term, they are "beta males".

If East Asians haven't accomplished as much as the test scores would suggest, I'd blame the conformity and nose to the grindstone mentality of their societies. I'd bet that East Asians living in predominately Western environments would be a lot more creative than those living in East Asia.

Westerners are more individualistic and tend to go in their own direction, which makes their societies more dynamic. However, unlike East Asians, they can't get their kids to sit down, behave themselves, and study for long durations of time. So they're not going to do as well on PISA or TIMSS, even if they have similar cognitive abilities.

Places like Japan, Korea, and Taiwan are pretty decent to live in and have strong economies. Sort of boring and the academic culture is too intense, but otherwise not bad at all. So I think Asians have a lot to be happy about and probably don't sit around pining to be Westerners.

Personality is an important factor, beyond IQ, to explain differences among groups.

Anonymous said...

http://super-economy.blogspot.com/2010/12/amazing-truth-about-pisa-scores-usa.html

European immigrants removed, and Math and Science added.

The Wobbly Guy said...

I've heard my local education officials comment recently that Japan's education system is in need of a drastic overhaul - they're putting almost the entire academic load on the private cram schools who are devoted to sheer rote learning stuff.

Perhaps given the discrepancy between IQ scores and achievement scores, we can use that difference as a way to assess the merits of each education system.

I have no doubts that if you take a pair of identical twins, and then put one in a US/Western system, and the other in a East Asian system, the achievement score will be different. But, their life achievements may still end up the same... and won't that be interesting!

Anonymous said...

There's a lot of unwarranted self-congratulation in assuming East Asians kids are grinds compared to individualistic and creative American/European kids.

When I see the youth culture of 15 year olds in Britain I don't see creative brilliance. Instead there's a Clockwork Orange descent into buffoonery and anti-social behavior.

White children in Britain, probably more than in any European country, are just becoming standard issue fuck ups.

Anonymous said...

East Asian kids are grinds, but grinds are better than clownish and imbecilic buffoons. You give me a kid that's motivated and respectful, and I'll take him over the wasted talent. I'd take East Asians over Britons any day.

I don't think anyone can deny that the East Asians have their house in order much better than any other ethnicity. A lot of time is spent documenting the failings of NAMs, but a lot of white European countries have pretty abysmal academic cultures. Italy, Greece, southeastern Europe, eastern Europe, and Britain are more known for their thugs and delinquents than any creative endeavor.

I think whites in Finland or the Netherlands or Canada are a better representation of intelligent and extracirricular oriented kids that just don't have much of an academic culture.

By the way, Indian and Pakistani kids seem to academically do really well in England and are highly represented in higher education. What's the explanation there? It'd seem to me that such good schooling results are not consistent with the 80 IQ given for South Asians...

Anonymous said...

Back in the day, back when "made in America" was synonymous with "quality," before all American manufacturing got offshored, the synonym for "crap" was "made in Taiwan."

Are the Han Chinese dull-witted or do they just not give a crap?


And before that, the synonym for "crap" was "made in Japan".

Taiwanese products aren't crappy anymore. If you asked a computer geek to build you a high quality computer today, he would make damn sure that many of the components are Taiwanese brands (Asus and Gigabyte come to mind, among others).

Anonymous said...

There's a lot of unwarranted self-congratulation in assuming East Asians kids are grinds compared to individualistic and creative American/European kids.

No, that's what the historical record indicates. If those East Asian kids ever manage to translate their success on academic tests into actual real world accomplishment (such as coming up with the calculus, or inventing the airplane, or the internal combustion engine, or antibiotics) then all this triumphalist East Asian chest-beating will be at least slightly justified. Nowhere near fully justified, but at least slightly.

We're nowhere near that point at present though.