5 ms·

Comparing a heterogeneous group to homogeneous groups, in three charts. Broken out, America doesn't look so bad: http://www.vdare.com/articles/pisa-scores-show

by nerdo 13y ago

Comparing a heterogeneous group to homogeneous groups, in three charts. Broken out, America doesn't look so bad: http://www.vdare.com/articles/pisa-scores-show-demography-is-destiny-in-education-too-but-washington-doesnt-want-you-to-k http://www.vdare.com/articles/pisa-scores-show-demography-is...

mikeash 13y ago
And why is that a better way to examine the numbers than the one in the article, exactly?
- jamesaguilar 13y ago
  Under the assumption that you're interested in knowing the cause of the problem, gaining insight into the factors that lead to or indicate the problem can aid in understanding. To speak in specifics rather than generalities: in this case, we see that White and Asian children do not seem to be having problems compared to their peers in other countries. Thus, we can assign a lower probability estimate to the hypothesis that the problem is the American educational system writ large. It seems to be a problem localized in certain parts of the system and population.
  - mikeash 13y ago
    Sure, but does this give any insight? Examining one factor out of the thousands that could potentially affect this does not strike me as in any way useful, especially since it happens to be a factor that's certain to generate lots of emotion and very little rational discussion. For example, it strikes me as unlikely that these factors are actually caused by race directly. I don't buy into the PC idea that everybody is exactly, precisely the same, but the differences simply aren't all that large. What seems much more likely is that it's related to poverty. Poverty and race are highly correlated in the US for various historical reasons, and I'd wager that the charts would show just as much, if not more, link between poverty and scores as between race and scores. But we can't say for certain without a proper analysis of all potential confounding factors. Picking a single one and pretending it's automatically important is the wrong way to do it.
    - ars 13y ago
      > it strikes me as unlikely that these factors are actually caused by race directly ... but the differences simply aren't all that large. These are things you want to believe, but they are not true. The reality is that race plays a huge role in intelligence. It's totally unfair, but it is what it is. Ignoring it, and believing otherwise does not make it go away, or make it better. It makes it worse. > Poverty and race are highly correlated in the US for various historical reasons Poverty is also highly correlated with intelligence. More correlated than it is to race. > link between poverty and scores as between race and scores. Of course, because poverty is a proxy to intelligence. > Picking a single one and pretending it's automatically important is the wrong way to do it. Pretending that the main one is automatically not-important because it's unfair, and/or huts your world view, is also the wrong way to do it.
      - mikeash 13y ago
        "The reality is that race plays a huge role in intelligence." I'm going to need some kind of cite for that before we go any further, you know.
        HelloMcFly 13y ago
        I've never seen this claim convincingly supported in a way that partials out the effect of socioeconomic status, parental involvement, and geography. It's an easy claim to support if you just look at straight averages. I'm sure there is information I'm not familiar with, but I'm unimpressed with the "citations" that have followed this claim in other conversations.
        thowawayanother 13y ago
        In the United States, low income whites perform better than high income blacks on the SAT - it's not a fact I suggest you go around citing at parties[1] " Blacks from families with incomes of more than $100,000 had a mean SAT score that was 85 points below the mean score for whites from all income levels, 139 points below the mean score of whites from families at the same income level, and 10 points below the average score of white students from families whose income was less than $10,000." [1] http://www.jbhe.com/features/49_college_admissions-test.html http://www.jbhe.com/features/49_college_admissions-test.html
  - ryandrake 13y ago
    You could just as easily plot USA-RICH and USA-POOR and conclude that wealth/income "lead to or indicate the problem". Or USA-NORTH, USA-SOUTH, USA-EAST and USA-WEST and conclude that geography "lead to or indicate the problem". I'm not sure how you can draw a causal relationship here.
    - jamesaguilar 13y ago
      I would be happy to see that done. I'm not making a claim about the cause of the problem except to observe that it doesn't seem to be affecting everyone. What I'd really love to see is an application that allows you to tweak the various group-bys (I'd like to try grouping by cost-of-living-adjusted log income brackets, personally, because my current estimate of that being the most related metric is highest. Or I'd like to see what happens if you take groups of kids with the same ratial and income backgrounds from various countries. Hold some variables constant, basically.). I am only saying exactly what I said: it doesn't seem that the US educational system, in its entirety, is significantly defective relative to other countries. Some combination of the attributes of certain American populations and/or the system's effect on them is causing those populations to perform much worse than kids in European and Asian countries. This in no way spells doom for our entire country as the article argues.
  - PeterisP 13y ago
    Let's be honest, race isn't an independent factor - at least in USA, it's highly correlated with family income and education level. If you're comparing white kids in USA with all kids in Germany, then you're comparing apples with oranges - if you're putting a filter that excludes a large portion of poor families in USA, then you should exclude the same portion of poor families from the German results. And if some other system is handling all social groups decently well, but USA fails with "certain parts of the system and population" as you say; then it's definitely a problem with "American educational system writ large".
- temphn 13y ago
  http://en.wikipedia.org/wiki/Thoughtcrime#Crimestop http://en.wikipedia.org/wiki/Thoughtcrime#Crimestop Crimestop means the faculty of stopping short, as though by instinct, at the threshold of any dangerous thought. It includes the power of not grasping analogies, of failing to perceive logical errors, of misunderstanding the simplest arguments if they are inimical to Ingsoc, and of being bored or repelled by any train of thought which is capable of leading in a heretical direction. The mind should develop a blind spot whenever a dangerous thought presented itself. The process should be automatic, instinctive. Crimestop, they called it in Newspeak. He set to work to exercise himself in crimestop. He presented himself with propositions -- 'the Party says the earth is flat', 'the party says that ice is heavier than water' -- and trained himself in not seeing or not understanding the arguments that contradicted them.
  - mikeash 13y ago
    What is the relevance of this quote?
  - ehmish 13y ago
    I take it you're getting at people avoiding thinking about the thought that race might be correlated with intelligence, lest they become an awful person. And it's a fair criticism. At the same time however, it has been noted that people have a subjective bias against people not of their own race.
  - wonderzombie 13y ago
    Race is absolutely a factor, but only insofar as white people enslaved black people, then legislated them into second class citizenship. Post-1960s, I guess we're on to blaming them (hence the stupid & lazy stereotype) because we think they should've gotten over it & recovered by now.
    - thowawayanother 13y ago
      I wonder what terrible things Asians did to whites such that they earn higher income, commit fewer crimes, and do better in schools than whites while living in historically white countries.
- thowawayanother 13y ago
  If the United States is becoming a Latin American country, why would you expect it to have macrostatistics comparable to Finland? It would be interesting to see international test scores from 1960, when the US was ~90% people of European descent. I would bet the US would compare closer to Europe on those. The demographic changes in the US over the last 50 years have been huge, and you can't ignore them in international comparisons.
jcampbell1 13y ago
I'd like to add that the Shanghai numbers are highly skewed. Half the students in Shanghai live there illegally, and attend informal schools. The Shanghai numbers represent the performance of the wealthiest 50%.

by nerdo 13y ago

mikeash 13y ago

And why is that a better way to examine the numbers than the one in the article, exactly?

jamesaguilar 13y ago
Under the assumption that you're interested in knowing the cause of the problem, gaining insight into the factors that lead to or indicate the problem can aid in understanding. To speak in specifics rather than generalities: in this case, we see that White and Asian children do not seem to be having problems compared to their peers in other countries. Thus, we can assign a lower probability estimate to the hypothesis that the problem is the American educational system writ large. It seems to be a problem localized in certain parts of the system and population.
- mikeash 13y ago
  Sure, but does this give any insight? Examining one factor out of the thousands that could potentially affect this does not strike me as in any way useful, especially since it happens to be a factor that's certain to generate lots of emotion and very little rational discussion. For example, it strikes me as unlikely that these factors are actually caused by race directly. I don't buy into the PC idea that everybody is exactly, precisely the same, but the differences simply aren't all that large. What seems much more likely is that it's related to poverty. Poverty and race are highly correlated in the US for various historical reasons, and I'd wager that the charts would show just as much, if not more, link between poverty and scores as between race and scores. But we can't say for certain without a proper analysis of all potential confounding factors. Picking a single one and pretending it's automatically important is the wrong way to do it.
  - ars 13y ago
    > it strikes me as unlikely that these factors are actually caused by race directly ... but the differences simply aren't all that large. These are things you want to believe, but they are not true. The reality is that race plays a huge role in intelligence. It's totally unfair, but it is what it is. Ignoring it, and believing otherwise does not make it go away, or make it better. It makes it worse. > Poverty and race are highly correlated in the US for various historical reasons Poverty is also highly correlated with intelligence. More correlated than it is to race. > link between poverty and scores as between race and scores. Of course, because poverty is a proxy to intelligence. > Picking a single one and pretending it's automatically important is the wrong way to do it. Pretending that the main one is automatically not-important because it's unfair, and/or huts your world view, is also the wrong way to do it.
    - mikeash 13y ago
      "The reality is that race plays a huge role in intelligence." I'm going to need some kind of cite for that before we go any further, you know.
      - HelloMcFly 13y ago
        I've never seen this claim convincingly supported in a way that partials out the effect of socioeconomic status, parental involvement, and geography. It's an easy claim to support if you just look at straight averages. I'm sure there is information I'm not familiar with, but I'm unimpressed with the "citations" that have followed this claim in other conversations.
      - thowawayanother 13y ago
        In the United States, low income whites perform better than high income blacks on the SAT - it's not a fact I suggest you go around citing at parties[1] " Blacks from families with incomes of more than $100,000 had a mean SAT score that was 85 points below the mean score for whites from all income levels, 139 points below the mean score of whites from families at the same income level, and 10 points below the average score of white students from families whose income was less than $10,000." [1] http://www.jbhe.com/features/49_college_admissions-test.html http://www.jbhe.com/features/49_college_admissions-test.html
- ryandrake 13y ago
  You could just as easily plot USA-RICH and USA-POOR and conclude that wealth/income "lead to or indicate the problem". Or USA-NORTH, USA-SOUTH, USA-EAST and USA-WEST and conclude that geography "lead to or indicate the problem". I'm not sure how you can draw a causal relationship here.
  - jamesaguilar 13y ago
    I would be happy to see that done. I'm not making a claim about the cause of the problem except to observe that it doesn't seem to be affecting everyone. What I'd really love to see is an application that allows you to tweak the various group-bys (I'd like to try grouping by cost-of-living-adjusted log income brackets, personally, because my current estimate of that being the most related metric is highest. Or I'd like to see what happens if you take groups of kids with the same ratial and income backgrounds from various countries. Hold some variables constant, basically.). I am only saying exactly what I said: it doesn't seem that the US educational system, in its entirety, is significantly defective relative to other countries. Some combination of the attributes of certain American populations and/or the system's effect on them is causing those populations to perform much worse than kids in European and Asian countries. This in no way spells doom for our entire country as the article argues.
- PeterisP 13y ago
  Let's be honest, race isn't an independent factor - at least in USA, it's highly correlated with family income and education level. If you're comparing white kids in USA with all kids in Germany, then you're comparing apples with oranges - if you're putting a filter that excludes a large portion of poor families in USA, then you should exclude the same portion of poor families from the German results. And if some other system is handling all social groups decently well, but USA fails with "certain parts of the system and population" as you say; then it's definitely a problem with "American educational system writ large".
temphn 13y ago
http://en.wikipedia.org/wiki/Thoughtcrime#Crimestop http://en.wikipedia.org/wiki/Thoughtcrime#Crimestop Crimestop means the faculty of stopping short, as though by instinct, at the threshold of any dangerous thought. It includes the power of not grasping analogies, of failing to perceive logical errors, of misunderstanding the simplest arguments if they are inimical to Ingsoc, and of being bored or repelled by any train of thought which is capable of leading in a heretical direction. The mind should develop a blind spot whenever a dangerous thought presented itself. The process should be automatic, instinctive. Crimestop, they called it in Newspeak. He set to work to exercise himself in crimestop. He presented himself with propositions -- 'the Party says the earth is flat', 'the party says that ice is heavier than water' -- and trained himself in not seeing or not understanding the arguments that contradicted them.
- mikeash 13y ago
  What is the relevance of this quote?
- ehmish 13y ago
  I take it you're getting at people avoiding thinking about the thought that race might be correlated with intelligence, lest they become an awful person. And it's a fair criticism. At the same time however, it has been noted that people have a subjective bias against people not of their own race.
- wonderzombie 13y ago
  Race is absolutely a factor, but only insofar as white people enslaved black people, then legislated them into second class citizenship. Post-1960s, I guess we're on to blaming them (hence the stupid & lazy stereotype) because we think they should've gotten over it & recovered by now.
  - thowawayanother 13y ago
    I wonder what terrible things Asians did to whites such that they earn higher income, commit fewer crimes, and do better in schools than whites while living in historically white countries.
thowawayanother 13y ago
If the United States is becoming a Latin American country, why would you expect it to have macrostatistics comparable to Finland? It would be interesting to see international test scores from 1960, when the US was ~90% people of European descent. I would bet the US would compare closer to Europe on those. The demographic changes in the US over the last 50 years have been huge, and you can't ignore them in international comparisons.

jcampbell1 13y ago

I'd like to add that the Shanghai numbers are highly skewed. Half the students in Shanghai live there illegally, and attend informal schools. The Shanghai numbers represent the performance of the wealthiest 50%.