« Neat business card designs | Main | Suzuki House »
Measuring Up
The subtitle of this excellent book, by Daniel Koretz, is What Educational Testing Really Tells Us. Here is one excerpt:
The distressingly large achievement differences among racial/ethnic groups and socioeconomic groups in the United States lead many people to assume that American students must vary more in educational performance than others. Some observers have even said that the horse race -- simple comparisons of mean scores among countries -- is misleading for this reason. The international studies address this question, albeit with one caveat: the estimation of variability in the international surveys is much weaker than the estimation of averages.
...We are limited to more general conclusions, along the lines of "the standard deviations in the United States and Japan are quite similar." Which they are. In fact, the variability of student performance is fairly similar across most countries, regardless of size, culture, economic development, and average student performance.
I was shocked to read this but the book is highly reputable and persuasive.
Posted by Tyler Cowen on May 21, 2008 at 12:39 PM in Education | Permalink
Comments
Does he mean that people who do well in school do well in life both places and that the gap in test scores are simlar to gap in icomes or that the gaps in test scores overall are simlar in the USA and Japan?
BTW Maybe we should say that the country in which people have the highest standard of living (excepting the oil rich) is the country with the best education. IMO schools are a very small part of education and the tests do not test important things.
Posted by: Floccina at May 21, 2008 2:31:03 PM
Maybe we should say that the country in which people have the highest standard of living (excepting the oil rich) is the country with the best education.
Information is more important. An illiterate plumber in the U.S. was richer than a highly educated Soviet planner because he had better information.
As to the book, wouldn't any large group of humans show similar variation in educational achievement, since the educational system is a sorting mechanism for society? If the American system were transferred to Japan and the Japanese system to the U.S., for instance, might the variance widen/narrow?
Posted by: 8 at May 21, 2008 2:46:08 PM
There are all sorts of reasons why comparing the USA with, at least, European countries is often an exercise in futility. Most notably, the early selection process that that determines your life by bifurcating education outcomes at around age 12 --- you have good grades you go to gymnasium (college preparatory) or you have bad grades you go to a technical school and become a plumber.
One should also not how the lack of 'diversity' is inversely related to educational success.
Posted by: Varangy at May 21, 2008 4:22:45 PM
8,
My understanding is that he is talking about the results of internationally given tests(Which are essentially the same, except for translation).
The distribution of these scores has a variance(The sum of the distance from the mean squared), and apparently, this variance is roughly the same in any group of people. Mind you, the actual average scores vary tremendously. But the percentage of kids who score 30% higher than the mean is constant across societies.
I guess I don't find this particularly surprising. Teaching methods are roughly similar around the world, the only variance is in resources.
As resources go up, there is a pretty strong public choice incentive for it to be distributed fairly equally among everyone, leaving the net effect on variance to be 0. Of course, the richest kids get vastly disproportionate funds, but they are too rare to really effect the numbers.
If there existed nations that had drastically different teaching styles(Something other than the traditional teacher-class relationship), it might look a little different.
Posted by: David Shor at May 21, 2008 4:37:59 PM
The recurring PISA exams --- given to 15 year-olds across 30 countries or so --- show in the latest 2008 report for the latest PISA that the variability of US student performance is in line generally with most other countries. In science, our students are a tad below average (above quite a few EU countries, below even more EU and East Asian ones). In math, worse yet. In reading --- (not clear from the charts (easy to read otherwise) if this is for scientific reading or general reading --- a no-show for reasons unclear from the PISA report.
Click here: http://earged.meb.gov.tr/yeni/duyurular/13_04_2008/dosyalar/ingilizce/science_competencies_for_tomorrows_world.swf
Michael (Gordon), AKA, the buggy professor: http://www.thebuggyprofessor.org
Posted by: michael gordon at May 21, 2008 4:44:20 PM
Let's look at the latest available PISA scores, the 2006 Science Literacy test. Among the 30 OECD countries, America came in 21st of 30 with a score of 483, below the average of 500. However, non-Hispanic white American students scored 523, which would have come in 7th out of 30.
Of course, other OECD countries, such as France (495) have similar factors at work. The countries at the top of the OECD list tend to be ones either with very little immigration (Finland was #1 by far with a 564 and Japan was #3) or a selective system of immigration (Canada was #2).
The largest source of immigrants to America, Mexico, was last out of the 30 OECD countries with a 410. (I suspect though that that's an improvement over where it would have stood a decade or two -- recent educational reforms in Mexico may be doing some good.) Turkey, which is probably the largest single source of immigrants to Western Europe, was next to last among OECD countries with a 424.
Among non-OECD countries, the top three were Hong Kong, Taiwan, and Finland-like Estonia. The bottom two were Qatar and Kyrgyz. No African countries were tested.
For details, see:
http://nces.ed.gov/pubs2008/2008016.pdf
Posted by: Steve Sailer at May 21, 2008 5:37:03 PM
Excuse me, the overall US score was 489 rather than 483.
Posted by: Steve Sailer at May 21, 2008 5:40:40 PM
American achievement differences were not significantly different from the average, but were quite a bit greater than high-achieving countries like Finland and Holland.
And American achievements were, as n oted, 'a tad below average', which is essentially, the middle of the pack, and well below high-achieving countries.
And people who explain the results by saying that other countries are less diverse will be unable to explain why Canada, which is very diverse, scored much better than the U.S.
Most people looking at the results have concluded that the route to better educational outcomes lies in the path the higher-achieving countries have all taken: greater attention to social and economic equity, higher pay and status for teachers, no separation of students into 'elite' and 'other' categories, limited testing, and greater emphasis on autonomy for individual schools.
Exactly the opposite of the approach advocated by the American right (are they not thoroughly discredited yet?).
Posted by: Stephen Downes at May 21, 2008 5:41:04 PM
Here's a simple way to look at variance, from the official PISA 2006 Science Literacy report:
"When comparing the performance of the highest
achieving students—those at the 90th percentile—
there was no measurable difference between the
average score of U.S. students (628) compared to
the OECD average (622) ...
"At the other end of the distribution, among low achieving students at the 10th percentile, U.S.
students scored lower (349) than the OECD average
(375) on the combined science literacy scale."
PISA ranks students on a seven level scale from Below Level 1 (don't ask) to Level 6 (Jimmy Neutron). Out of the 30 OECD countries, the US has a higher percentage (25%) at the two worst levels than all but Turkey and Mexico. In contrast, Finland has only 5% at those abysmal levels, and Japan only 12%.
See Figure 5 in:
http://nces.ed.gov/pubs2008/2008016.pdf
Posted by: Steve Sailer at May 21, 2008 5:50:09 PM
Perhaps grade inflation enters into the picture: in the US, students usually get at least a passing grade just for breathing. Another common practice is to adjust grades to fit a bell curve, so practically by definition the standard deviations will match what you'd expect.
Test designers create tests that produce a "reasonable" spectrum of test scores: any test that produced highly lopsided or narrowly distributed results would be rewritten or replaced. In any real setting, tests are adjusted to match students' capabilities: the goal after all is to discover where each student stands relative to his or her peers, not to discover anything about statistical distributions. The statistical distributions are more or less chosen in advance, like zooming a lens in or out to fill a camera's field of view.
The excerpt sounds like stating a tautology rather than some unexpected insight.
Posted by: at May 21, 2008 5:53:29 PM
Stephen Downes asserts:
"And people who explain the results by saying that other countries are less diverse will be unable to explain why Canada, which is very diverse, scored much better than the U.S."
The word "diversity" can mean a lot of things. What really matters is that Canada has higher median human capital than America. Canada has very little illegal immigration and thus practically no Mexicans, and is only about 2% black. It's system of legal immigration is explicitly designed to benefit current Canadian citizens by carefully selecting those applicants with the highest human capital. I, for example, took the Canadian immigration online assessment in 2001 for an article I was writing and failed to score high enough to qualify for an interview with a Canadian immigration official. (Their opinion was that they had plenty of journalists already, thank you very much, don't call us, we'll call you.)
Here's the executive summary of the latest PISA report from the federal National Center for Educational Statistics on U.S. performance:
"On the combined science literacy scale, Black
(non-Hispanic) students (409) and Hispanic
students (439) scored lower, on average, than
White (non-Hispanic) students (523), Asian (non-
Hispanic) students (499), and students of more
than one race (non-Hispanic) (501)."
[The OECD average is set to 500.]
So, African-Americans score the same in science as Mexicans do in Mexico. Hispanics in America due considerably better than Mexicans in Mexico, but much worse than non-Hispanic whites and Asians.
By the way, educational performance by Mexican-Americans, while it improves considerably between the immigrant generation and the first generation not born in America, does _not_ improve in subsequent generations, leaving Mexican-Americans permanently well below the non-Hispanic white average. See the landmark 2008 book "Generations of Exclusion" by Telles et al of the UCLA Chicano Studies Center for a brilliant study comparing Mexican-Americans in 1965 and their descendants in 2000.
Posted by: Steve Sailer at May 21, 2008 6:08:01 PM
Not surprisingly, it turns out that, contrary to Tyler's posting, the U.S. does have an above average standard deviation on the 2006 PISA exam of science literacy among 15 year olds. The OECD average standard deviation was 95 points (on a 0 to 1000 scale where 500 is average), and the U.S. has the third highest, 106, just behind the UK and New Zealand's 107. Japan is 100.
The lowest standard deviations are in Mexioo and Turkey, due to their having so few high scorers and such low average scores. Interestingly, though, superstar and super-homogenous Finland has the third lowest standard deviation at 86.
See Table C4 on p. 45 of:
http://nces.ed.gov/pubs2008/2008016.pdf
Posted by: Steve Sailer at May 21, 2008 6:30:13 PM
If you don't want to buy the "Generations of Exclusion" book that Steve Sailer mentions, you can still read the informative UCLA press report:
"Mexican American integration slow, education stalled, study finds"
http://newsroom.ucla.edu/portal/ucla/ucla-study-of-four-generations-46372.aspx
excerpt: "unlike the descendants of European immigrants to the United States, Mexican Americans have not fully integrated by the third and fourth generation."
Posted by: scottynx at May 21, 2008 6:30:56 PM
The word "diversity" can mean a lot of things. What really matters is that Canada has higher median human capital than America. Canada has very little illegal immigration and thus practically no Mexicans, and is only about 2% black. It's system of legal immigration is explicitly designed to benefit current Canadian citizens by carefully selecting those applicants with the highest human capital. I, for example, took the Canadian immigration online assessment in 2001 for an article I was writing and failed to score high enough to qualify for an interview with a Canadian immigration official. (Their opinion was that they had plenty of journalists already, thank you very much, don't call us, we'll call you.)
Here's the executive summary of the latest PISA report from the federal National Center for Educational Statistics on U.S. performance:
"On the combined science literacy scale, Black
(non-Hispanic) students (409) and Hispanic
students (439) scored lower, on average, than
White (non-Hispanic) students (523), Asian (non-
Hispanic) students (499), and students of more
than one race (non-Hispanic) (501)."
[The OECD average is set to 500.] --- Steve Sailer
..............
That's a good post, Steve, full of solid information . . . along with the useful link to the NCES report.
It does lead to an important query: why do libertarians and generally free-market-enthusiasts on one side and politically correct radicals of all sorts on the left side of the political spectrum (except for African-American studies profs) continue to support a policy of either unlimited illegal immigration mainly out of Mexico and Central America or an amnesty with a promise by the government once more for legalizing their status?
The reasons the two sides of the ideological spectrum support such illegal immigration --- never mind lopsided legal immigration out of the poorer Latin American countries --- do differ, obviously. No matter. Both sides, along with powerful interest groups, continue to favor policies that are creating the basis of a new, ever larger ethnic-based underclass. As for educational performance, the response from the radical and liberal left is the same mantra as in the past: not enough money spent in the poorer performing school areas. Yet the only study ever produced by the US Dept of Education (1990, I believe) showed an insignificant difference in, say, the amount of money spent on wholly or overwhelmingly white-attended schools and wholly or overwhelmingly black-attended ones. (There were further sub-divisions according to mixed race/ethnic schools, with similar correlations found.) And the experiments carried out --- lavish sums spent in Kansas City, Mo in the 1990s --- showed no noticeable improvement in school performance whatsoever.
In the L.A. public school system, 60% or so of Hispanic students do not graduate from high school . . . a figure higher than the equivalent for black students (about 55%). At the same time, gang fights and other racially inspired attacks on a large scale are occurring in black-Hispanic schools . . . as happened recently in L.A., where nearly a thousand students were involved (I believe) in the melee. That happens to be a school where only 3% of the small number of graduates go on to college.
...............
Enter, as just indicated, the strains on the social fabric of local communities thanks to higher and higher numbers of illegal (and to an extent legal) Hispanic immigrants. There's the huge resource-strain in L.A.'s public clinics, hospitals, and ER everywhere. Then there's violent crime. In Santa Barbara where I live, Hispanic gangs have mushroomed in size and violence, with three killings in the last year . . . not to mention other killings and armed robberies that did not exist 15 years ago. And of course there is the continued increase in the supply of poorly educated workers who, while keeping the wages down in restaurants, hotels, sweat-labor shops, and farming, have --- with indifference by libertarians (who hail it as a big gain in labor efficiency --- left other potential workers in those industries with little interest in pursuing jobs there, even though they might have lost better paying jobs owing to factors way beyond their control: such as massive technological changes and the ever greater pace of globalizing influences.
How much exactly would wages rise in those industries if the ever larger pool of poorer immigrants workers was shut off? And would the impact on the US average living standard --- or on well-to-do Americans who travel and stay in hotels and eat out in restaurants beyond the level of MacDonald's --- even remotely approach the shock, say, of ever higher gas-prices.
............
Yet we're told by the radical and liberal left: it's racism to start shifting toward a Canadian based immigration system, with those who have lost better paying jobs taking refuge, a la Obama's views, in guns, cars, and religion. And on the right libertarian and free-market side: hey, that's life, markets always produce optimal results unless interfered with by government regulations or distributional policies, and go get re-trained, re-educated, and pick up and move to a distant community, guys!
I wonder how many tenured professors in their late 40's, 50's, and early 60's would like to emulate the advice?
............
Small wonder, on the conservative side of the spectrum, the Republican Party is increasingly being abandoned as an out-of-date, unresponsive, uncaring coalition of libetarians, well-to-do evangelicals (middle class people generally), and free-market enthusiasts.
Believe it or not, neo-conservativism was a tag originally applied to former liberals who --- alarmed by the trends of the late 1960's and LBJ's new welfare policies --- formed the most influential policy-oriented journal in US history, THE PUBLIC INTEREST. And what distinguished them from paleo-conservatives and libertarians was their emphasis on the need for a decent if limited welfare-state and on cultural and community interests . . . all at odds with the libertarian and free-market stress on self-interested, atomistically viewed individuals pursuing their interests within the framework of a market. (Only later, in the 1980s, did a journal spinoff THE NATIONAL INTEREST emerge as part of the Reagan administration's shift toward a more aggressive form of a active and militarized diplomacy against the Soviet Union and its communist empire . . . practiced with remarkable flexibility by Ronald Reagan himself.)
--- Michael Gordon, AKA, the buggy professor: http://www.thebuggyprofessor.org
Posted by: michael gordon at May 21, 2008 7:05:16 PM
What converted me to the view that intelligence and human capital differences are likely intractable and quite probably genetically based? In large part, blog posts like this one, where people like Gordon and Sailer reference fact after fact after fact and go unanswered.
Posted by: Roberto at May 21, 2008 7:12:26 PM
Am I the only one who thinks these cross-country comparisons might be highly skewed by which people in which countries are likely to take these tests. E.g, I would guess that the average student in the UK who takes the International Baccalaureate is probably significantly better educated than the average student in the UK, purely because the IB is taken more commonly in private schools.
Posted by: John Faben at May 21, 2008 7:37:53 PM
So Roberto, why not come out and say that you are a racist if that is what you mean. If you mean something else, please explain.
Posted by: RobbL at May 21, 2008 8:04:33 PM
I don't think I'm a racist because I'd call racism a preference, all else held equal, for a particular race or races over others. Rather, I'd just call myself a rational person because I have looked at a body of evidence pertaining to a particular set of hypotheses and to the best of my ability updated my priors based on that evidence.
Posted by: roberto at May 21, 2008 9:41:32 PM
Realists believe the truth is better for humanity than lies, ignorance, or wishful thinking.
Lots of people, however, disagree.
Posted by: Steve Sailer at May 21, 2008 10:07:06 PM
Steven Pinker made an excellent point to me when I interviewed him in 2002:
?Q: Aren't we all better off if people believe that we are not constrained by our biology and so can achieve any future we choose?"
"A: People are surely better off with the truth. Oddly enough, everyone agrees with this when it comes to the arts. Sophisticated people sneer at feel-good comedies and saccharine romances in which everyone lives happily ever after. But when it comes to science, these same people say, "Give us schmaltz!" They expect the science of human beings to be a source of emotional uplift and inspirational sermonizing."
http://www.isteve.com/2002_QA_Steven_Pinker.htm
Posted by: Steve Sailer at May 21, 2008 10:09:36 PM
Steve Sailer,
Canada does not have a particularly selective system of immigration; among other things, it has the highest per-capita immigration rate of any large country.
Nearly half of immigrants can be considered "non-economic" migrants. There are some recent government proposals to prioritize the applications of more qualified immigrants, similar to Australia, but there is opposition and it's not clear whether anything will come of it.
There is also in fact a considerable amount of illegal immigration, in the form of bogus refugee applications. Refugee claims can take years to sort out, during which time claimants can legally work in Canada.
A large percentage of these bogus refugee claimants end up staying, even if their claim is rejected. Removal orders are only spottily enforced; sometimes even deported persons simply come back. People sometimes grumble about that sort of thing, but it isn't anything like the white-hot issue it is in the States or Europe. There is fairly broad support or at least tolerance of high levels of immigration across the political spectrum and within Canadian society, and not everyone cares whether you came in through the front door or the back door.
It's true that there are far fewer Mexicans in Canada than in the US, but that's just for obvious reasons of geography: why cross an extra border just to get to a colder place? Some of the Mexicans in Canada are there thanks to a guest-worker program. Interestingly, in 2005, Mexico ranked number one in the number of refugee claims by country of origin, followed by China.
I'm afraid the strict, selective mostly-all-legal Canadian immigration model that you hold up in contrast to the US policies that you oppose seems to be, well, mostly a product of your imagination.
Posted by: at May 21, 2008 10:12:51 PM
Steve,
It's a lot harder to estimate variance than it is to estimate mean. I doubt the differences in sample variance you mentioned are statistically significant, unless the sample sizes are truely huge.
As for everyone else: There may very well be a correlation between race and intelligence(Though, I'd point out that the IQ gap between blacks and whites does not exist in the UK or the Netherlands). But frankly, you are almost certainly a racist if you bring it up.
Why? Because there are far better indicators of IQ that are not so politically charged. Birth weight, childhood nutrition, and amount of stimulation in early development correlate much stronger with race then race does.
For these people, it is not about uncovering the truth. It is about confirming their dark world views.
Instead, I'll say this: A good deal of IQ depends on environmental factors during early development. Does this explain all inter-racial IQ variance? Maybe. Instead, keep in mind that a century ago, Jews and Asians had lower IQs than the norm.
Posted by: David Shor at May 22, 2008 12:07:06 AM
John Faben said: "Am I the only one who thinks these cross-country comparisons might be highly skewed by which people in which countries are likely to take these tests. E.g, I would guess that the average student in the UK who takes the International Baccalaureate is probably significantly better educated than the average student in the UK, purely because the IB is taken more commonly in private schools."
PISA tests are given to a sample of all 15 year olds in all types of schools, so there is noeffect of the type you conjecture. See http://www.oecd.org/pages/0,3417,en_32252351_32235907_1_1_1_1_1,00.html and links
Posted by: Peter Whiteford at May 22, 2008 6:11:31 AM
As a native Los Angeleno, I'm always amused by how people, especially in the DC and NYC areas, draw analogies to Ellis Island immigrant groups to predict the future performance of Mexican illegal immigrants and their descendants, as if there had never been any people of Mexican background in the U.S. until a few years ago. In reality, there were one million Hispanics in the U.S. according to the 1920 Census.
This new book by Edward E. Telles and Vilma Ortiz of the UCLA Chicano Studies Center "Generations of Exclusion: Mexican Americans, Assimilation, and Race" provides the definitive answer on Mexican-American progress. They found the questionnaires from a big study of Mexican Americans done in LA and San Antonio in 1965, tracked down the respondents, then re-interviewed them and their children in 2000. What they found was solid improvements in educational attainment from the first (immigrant) generation to the second (American-born) generation, but then, unlike Ellis Island immigrants, no further improvements in subsequent generations. On p. 133, Telles and Ortiz write:
"Rather than improvements in education in subsequent generations-since-immigration, as assimilation theory predicts, we find quite the opposite. ... Moreover, our study shows similarly poor educational outcomes for the fifth generation so far ..."
The fifth generation!
They argue:
"America's public schools have failed most Mexican Americans, contrary to what they did for European Americans."
Or, perhaps, Mexican Americans have tended to fail America's public schools. But that's not the important point here. What is important is that all the pundits' fantasies about how Mexican Americans are, on average, going to turn into the new Italian Americans or even the new Jewish Americans Real Soon Now are flapdoodle. It's been tested unto the fifth generation and as anybody in the Southwest could have pointed out, it's not happening.
Posted by: Steve Sailer at May 22, 2008 6:14:29 AM
Most people looking at the results have concluded that the route to better educational outcomes lies in the path the higher-achieving countries have all taken: greater attention to social and economic equity, higher pay and status for teachers, no separation of students into 'elite' and 'other' categories, limited testing, and greater emphasis on autonomy for individual schools.
Who are these "most people", and have the higher-achieving countries actually taken the path you prefer? Many of the highest-achieving countries are East Asian, and as far as I know, the education system there is very traditional, with lots of rote learning and the like. Your recommendations describe the top-performing Finland better, although teachers are not well-paid in Finland, and testing and grading is very central to the Finnish system.
Posted by: JL at May 22, 2008 7:34:27 AM
Again, thanks to Steve Sailer and others who have posted in this thread. I did find two good, balanced sources that you might read if you have the time. They're chosen because, besides being balanced and not superficial analytically and data-wise, they're easy to read . . . intended for a mass public.
On US immigration and the debates pro- and -con, and various policy alternatives: http://www.nifi.org/stream_document.aspx?rID=10180&catID=6&itemID=10176&typeID=8
On Canada's immigration policy, part of a larger study --- well-done and informative --- that deals with the growing income inequality in Canada generally and in Ontario more specifically . . . Ontario, of course, the most populated Province along with Quebec.:
www.competeprosper.ca/download.php?file=WP10.pdf - Sep 10, 2007 The sections dealing with immigration are pp. 36ff, but the entire study is rewarding.
What I didn't find --- largely, perhaps, because I don't have the time available for such a study --- is an in-depth study of Canada's (or Australia's) skill-based immigrant policies, though a few other google-links I followed did indicate that they aren't as rigorously applied as I and other Americans might think. If anyone here can point to a solid, empirically based article or study that probes this and other matters discussed in this thread, he would be doing some of us a big favor.
-- Michael Gordon, AKA, the buggy professor: http://www.thebuggyprofessor.org
Posted by: michael gordon at May 22, 2008 10:59:28 AM
An interesting article in the Washington Post (from 2003) about the less-performing parts of the Japanese school system: Fast Times at Asakita High (via Craig Newmark)
Posted by: at May 22, 2008 11:17:14 AM
Over all good debate here
Steve Sailor wrote:
Not surprisingly, it turns out that, contrary to Tyler's posting, the U.S. does have an above average standard deviation
Steve Sailor how do you know that was what Tyler was saying? He could be, more profoundly, saying that poor people at x level below average income on average score x perfect below average on these tests and that the different ethnic groups scores differ similarly in both countries.
All of this indicates how small a factor schooling is on these test and therefore that the tests show ability more that schooling. IMO since this is true rather than trying to improve the schools to score better on those tests, we should try to teach more useful information (See Alfie Kohn and John Taylor Gatto). E.G. Much of the math that college bound students learn in high school is never used by more than a very, very few after we get out of school it is just a test. It could be the realm of math clubs and that time could be used to teach useful stuff.
Because of all this I recommend to students that they think of school as a long test and use any method (tutors, charm, asking lots of questions) aside from cheating to beat the system.
Posted by: Floccina at May 22, 2008 12:32:39 PM
Why not cheating? If you do recommend "charm"?
Posted by: greatzamfir at May 22, 2008 12:56:36 PM
A question I'd very much like the answer to is why participants in these sort of discussions that disagree with Steve tend to frequently and absurdly misspell his last name, despite its prominent display all over the page. For a commentariat that is as intelligent and educated as MR's is, it seems a pretty juvenile thing to do. But it seems to happen across sharply different blogs and fora. I wonder what the psychological motivation for that is - is it somehow ingrained?
Posted by: Billare at May 22, 2008 3:10:57 PM
He could be, more profoundly, saying that poor people at x level below average income on average score x perfect below average on these tests and that the different ethnic groups scores differ similarly in both countries.
Why is this profound?
Posted by: KDeRosa at May 22, 2008 10:55:02 PM
Billera sorry I am just a bad speller. Also Though I am not 100% with him I am more on Steve side in the debate than not.
KDeRosa it could mean that relative income has a large impact on school outcomes or that ability in school correlates very well with ability to earn even across different countries. It might also imply that we in the USA have a greater variance in income because our people vary more in ability.
Posted by: Floccina at May 23, 2008 9:14:22 AM
