6 ms·

Plenty of people recognize that the SAT is bias against poor kids https://www.theatlantic.com/education/archive/2014/07/why-poor-schools-cant-win-at-standardiz

by cmh89 5y ago

Plenty of people recognize that the SAT is bias against poor kids https://www.theatlantic.com/education/archive/2014/07/why-poor-schools-cant-win-at-standardized-testing/374287/ https://www.theatlantic.com/education/archive/2014/07/why-po... The reason that the racial component comes out is that Black children are over-represented in poor communities due to systemic racism

atweiden 5y ago
What JPKab said is entirely true. The SAT is known to correlate strongly with general intelligence, which is determined almost exclusively by hereditary factors outside a person’s control [1]. The data clearly suggests lower class persons of east asian descent will disproportionately suffer the consequences of this policy. [1]: https://www.nytimes.com/1994/10/16/books/what-is-intelligence-and-who-has-it.html https://www.nytimes.com/1994/10/16/books/what-is-intelligenc... [1]: https://www.nytimes3xbfgragh.onion/1994/10/16/books/what-is-intelligence-and-who-has-it.html https://www.nytimes3xbfgragh.onion/1994/10/16/books/what-is-...
temp8964 5y ago
What? The article you linked does not address SAT at all!
livinginfear 5y ago
SAT results correlate highly with IQ, and IQ correlates highly with success in life. Still, people engage in all kinds of postmodern mental gymnastics to explain why an intuitively obvious fact is somehow false.
- mrh0057 5y ago
  Your statement immediately has a problem. How do you define success? Then you have a second problem when you state a person taking one test is also good on a different test. Going much over 100 has little correlation with wealth which is what most studies use. Then you run into correlation doesn’t equal causation problem too.
  - livinginfear 5y ago
    > " How do you define success?" Surely you jest. It really doesn't matter how you define success, IQ is a valid predictor of pretty much any definition of success you can think of. Income, academic achievement, you name it.
    - chillwaves 5y ago
      Defining IQ is tricky to begin with. So is defining success. Do you really not see the issue?
      - setr 5y ago
        Under GP’s terms, success is trivial to define — pick a reasonable metric. IQ is also well defined — general intelligence is not, and whether IQ maps to general intelligence is questionable (because the latter is hard to define). But for this argument, you don’t need to define general intelligence, and really even IQ’s definition doesn’t matter as long it’s stable (define IQ as simply the outcome of an IQ test). Regardless of IQ’s definition, or the precise questions posed, it simply needs to (consistently) correlate with your selected definition of success to be useful. He’s also served the counter argument on a plate: pick a reasonable definition of success, and find little or no correlation to IQ. You only need to agree the definition is “reasonable”, and the study is not flawed.
  - inglor_cz 5y ago
    "How do you define success?" Let us look at the USA of today and its social ills. Quality of a random person's life is undeniably better if they manage: * to avoid going into jail/prison, * to always have a home, * to be drug-free, or at least drug addiction-free, * to be fit (= not obese), * to be employed or self-employed, * to finish their high school education at least. Those points might sound modest, but if someone can tick off all of them, they are already sorta successful, especially compared to someone who can only tick off one or two. A nation of people who could tick off 5 of 6 would probably be much happier than contemporary America. Do all those points correlate with IQ? I would guess yes, but IANA Social Scientist.
JPKab 5y ago
The confidence with which you assert "systemic racism" as being the key factor causing Black kids to be overrepresented in poor communities is fascinating. Historical racism, which was woven into law, is the actual reason for this. Historical factors are also why White southerners are overrepresented in poverty statistics compared to Whites in the midwest and New England. Historical racism isn't the same thing as the nebulous, impossible to define "systemic" racism, which basically means whatever the activist group pushing it wants it to mean. Typically, they look at any discrepancy in outcomes between two groups, ignore the complex multi-variate nature of said issue, and reduce it down to race, making the claim of "systemic" racism. The vast majority of CS grads from universities are males, but you never see this mentioned in claims of "systemic sexism" in the tech industry. Gender activism, like race activism, only goes in one direction, always. The same activist groups tend to ignore factors such as 50% of all murders in the US being perpetrated by, and victimizing, Black Americans. This kind of statistic is inconvenient for their narrative of police deliberately arresting people based on race, vs. the fact that they commit crimes disproportionately. Also, constantly cited is the discrepancy in crack penalties vs. cocaine penalties. This ignores the fact that crack and meth (a drug used predominantly by Whites) have equivalently harsh penalties, and also ignores the fact that these harsh penalties were demanded and campaigned for by Black politicians in the 90's, reacting to the destruction of their communities by the crack epidemic. There is a clear record of this, as evidenced by Joe Biden being a huge proponent of the Crime Bill. He and other Democrats were being urged to push the Crime Bill by his own party's Black politicians. Growing up in the rural south in a trailer park, the subculture of White people I belong to commit crimes at vastly higher rates than Whites in other parts of the country. They are arrested and shot by police at equally higher rates as well. Again, the low-resolution, activist driven narrative ignores all of this, treating White people in New York City and in Kentucky as being part of a homogenous group, instead of correctly identifying the vast cultural differences between the two going back to the completely different European cultures they are descended from (Dutch/wealthy English vs. dirt poor Ulster Scots emigrants).
- bookofsand 5y ago
  For a less US-centric perspective, regional differentials in development persist even for countries with 'homogenous' populations. Italy is a canonical example, with significant differences between the rich North and the poor South. By US ethnic standards, Northern and Southern Italians are basically indistinguishable clones, though Italians themselves may beg to differ. See also Germany, Belgium, Poland or Romania. https://commons.wikimedia.org/wiki/File:Map_of_Italian_regions_by_GDP_per_capita_in_euros_(2015)_alternative_colours.jpg https://commons.wikimedia.org/wiki/File:Map_of_Italian_regio... https://en.wikipedia.org/wiki/List_of_German_states_by_GRP_per_capita https://en.wikipedia.org/wiki/List_of_German_states_by_GRP_p... https://en.wikipedia.org/wiki/List_of_Belgian_provinces_by_GDP https://en.wikipedia.org/wiki/List_of_Belgian_provinces_by_G... https://en.wikipedia.org/wiki/Poland_A_and_B https://en.wikipedia.org/wiki/Poland_A_and_B https://commons.wikimedia.org/wiki/File:Romania_-_Nominal_GDP_per_Capita_by_County.png https://commons.wikimedia.org/wiki/File:Romania_-_Nominal_GD...
JPKab 5y ago
Question: Do you think that targeting people based on income level would be a superior way to address this issue? If it was based on income, rather than skin color, wouldn't it disproportionately help Black children while simultaneously not leaving other disadvantaged groups out in the cold?

by cmh89 5y ago

atweiden 5y ago

What JPKab said is entirely true. The SAT is known to correlate strongly with general intelligence, which is determined almost exclusively by hereditary factors outside a person’s control [1]. The data clearly suggests lower class persons of east asian descent will disproportionately suffer the consequences of this policy. [1]: https://www.nytimes.com/1994/10/16/books/what-is-intelligence-and-who-has-it.html https://www.nytimes.com/1994/10/16/books/what-is-intelligenc... [1]: https://www.nytimes3xbfgragh.onion/1994/10/16/books/what-is-intelligence-and-who-has-it.html https://www.nytimes3xbfgragh.onion/1994/10/16/books/what-is-...

temp8964 5y ago

What? The article you linked does not address SAT at all!

livinginfear 5y ago

SAT results correlate highly with IQ, and IQ correlates highly with success in life. Still, people engage in all kinds of postmodern mental gymnastics to explain why an intuitively obvious fact is somehow false.

mrh0057 5y ago
Your statement immediately has a problem. How do you define success? Then you have a second problem when you state a person taking one test is also good on a different test. Going much over 100 has little correlation with wealth which is what most studies use. Then you run into correlation doesn’t equal causation problem too.
- livinginfear 5y ago
  > " How do you define success?" Surely you jest. It really doesn't matter how you define success, IQ is a valid predictor of pretty much any definition of success you can think of. Income, academic achievement, you name it.
  - chillwaves 5y ago
    Defining IQ is tricky to begin with. So is defining success. Do you really not see the issue?
    - setr 5y ago
      Under GP’s terms, success is trivial to define — pick a reasonable metric. IQ is also well defined — general intelligence is not, and whether IQ maps to general intelligence is questionable (because the latter is hard to define). But for this argument, you don’t need to define general intelligence, and really even IQ’s definition doesn’t matter as long it’s stable (define IQ as simply the outcome of an IQ test). Regardless of IQ’s definition, or the precise questions posed, it simply needs to (consistently) correlate with your selected definition of success to be useful. He’s also served the counter argument on a plate: pick a reasonable definition of success, and find little or no correlation to IQ. You only need to agree the definition is “reasonable”, and the study is not flawed.
- inglor_cz 5y ago
  "How do you define success?" Let us look at the USA of today and its social ills. Quality of a random person's life is undeniably better if they manage: * to avoid going into jail/prison, * to always have a home, * to be drug-free, or at least drug addiction-free, * to be fit (= not obese), * to be employed or self-employed, * to finish their high school education at least. Those points might sound modest, but if someone can tick off all of them, they are already sorta successful, especially compared to someone who can only tick off one or two. A nation of people who could tick off 5 of 6 would probably be much happier than contemporary America. Do all those points correlate with IQ? I would guess yes, but IANA Social Scientist.

JPKab 5y ago

The confidence with which you assert "systemic racism" as being the key factor causing Black kids to be overrepresented in poor communities is fascinating. Historical racism, which was woven into law, is the actual reason for this. Historical factors are also why White southerners are overrepresented in poverty statistics compared to Whites in the midwest and New England. Historical racism isn't the same thing as the nebulous, impossible to define "systemic" racism, which basically means whatever the activist group pushing it wants it to mean. Typically, they look at any discrepancy in outcomes between two groups, ignore the complex multi-variate nature of said issue, and reduce it down to race, making the claim of "systemic" racism. The vast majority of CS grads from universities are males, but you never see this mentioned in claims of "systemic sexism" in the tech industry. Gender activism, like race activism, only goes in one direction, always. The same activist groups tend to ignore factors such as 50% of all murders in the US being perpetrated by, and victimizing, Black Americans. This kind of statistic is inconvenient for their narrative of police deliberately arresting people based on race, vs. the fact that they commit crimes disproportionately. Also, constantly cited is the discrepancy in crack penalties vs. cocaine penalties. This ignores the fact that crack and meth (a drug used predominantly by Whites) have equivalently harsh penalties, and also ignores the fact that these harsh penalties were demanded and campaigned for by Black politicians in the 90's, reacting to the destruction of their communities by the crack epidemic. There is a clear record of this, as evidenced by Joe Biden being a huge proponent of the Crime Bill. He and other Democrats were being urged to push the Crime Bill by his own party's Black politicians. Growing up in the rural south in a trailer park, the subculture of White people I belong to commit crimes at vastly higher rates than Whites in other parts of the country. They are arrested and shot by police at equally higher rates as well. Again, the low-resolution, activist driven narrative ignores all of this, treating White people in New York City and in Kentucky as being part of a homogenous group, instead of correctly identifying the vast cultural differences between the two going back to the completely different European cultures they are descended from (Dutch/wealthy English vs. dirt poor Ulster Scots emigrants).

bookofsand 5y ago
For a less US-centric perspective, regional differentials in development persist even for countries with 'homogenous' populations. Italy is a canonical example, with significant differences between the rich North and the poor South. By US ethnic standards, Northern and Southern Italians are basically indistinguishable clones, though Italians themselves may beg to differ. See also Germany, Belgium, Poland or Romania. https://commons.wikimedia.org/wiki/File:Map_of_Italian_regions_by_GDP_per_capita_in_euros_(2015)_alternative_colours.jpg https://commons.wikimedia.org/wiki/File:Map_of_Italian_regio... https://en.wikipedia.org/wiki/List_of_German_states_by_GRP_per_capita https://en.wikipedia.org/wiki/List_of_German_states_by_GRP_p... https://en.wikipedia.org/wiki/List_of_Belgian_provinces_by_GDP https://en.wikipedia.org/wiki/List_of_Belgian_provinces_by_G... https://en.wikipedia.org/wiki/Poland_A_and_B https://en.wikipedia.org/wiki/Poland_A_and_B https://commons.wikimedia.org/wiki/File:Romania_-_Nominal_GDP_per_Capita_by_County.png https://commons.wikimedia.org/wiki/File:Romania_-_Nominal_GD...

JPKab 5y ago

Question: Do you think that targeting people based on income level would be a superior way to address this issue? If it was based on income, rather than skin color, wouldn't it disproportionately help Black children while simultaneously not leaving other disadvantaged groups out in the cold?