6 ms·

Test scores are race-blind and fair. Test instructors do not give more scores to a test because they see it was completed by an Asian test-taker. It's great th

Test scores are race-blind and fair. Test instructors do not give more scores to a test because they see it was completed by an Asian test-taker. It's great that schools are looking for a better "mix" of ethnicities in the incoming class, but they should encourage the underrepresented ethnicities to do better, in tests or other race-blind admission standards. They should not put an admission "cap" based on the color of your skin

swsieber 8y ago
This. The end of the pipeline is not the correct place to change pipeline demographics, and in cases like this, is actually harmful.
- squiggleblaz 8y ago
  But where is the right place to change pipeline demographics? The problem is that certain groups are intergenerationally poor. One mechanism that is available to us to fix that is education. So this isn't the end of the pipeline, it's just the start of the next pipe. Where do you propose we intervene? Should we give a cushy job to the parents of every disadvantaged minority person as soon as they're born? Maybe we should give an education to the parents of every disadvantaged minority person before they're even the apple in the eye of said parents?
  - confiscate 8y ago
    You intervene when the student is still learning the material, before the test. You do not intervene after the test. At that point the learning process for that material is already over. The student has already "missed out" on the learning opportunity. You would just be giving a good grade to someone who doesn't know the material. Yes we need to help folks who are intergenerationally poor. But, you do that by helping their kids learn better and meet the testing bar. You are not helping them by lowering the bar for them. Giving the kids better grades even if they don't know the material, would actually hurt them.
    - ytpete 8y ago
      > You are not helping them by lowering the bar for them. But you are helping underperforming students by placing them into top schools, comingled with better students - lots of research shows that. So lowering the bar for admission really does help. I think the open question is how that balances against the cost of having fewer slots available at those top schools for the very best students.
      - confiscate 8y ago
        I don't know what research you are pointing to. But I am pretty sure that if, say, a student can't pass a multiplication test, but you give that student an A+ anyways, and then let that student into a "top calculus class", that's going to hurt the student more than it helps. Ya sure the student gets to mingle with better students. But what you would usually get is, the student would end up being someone who mentions a lot of buzz words like "fractal integrals" just to fit in, but have zero idea what is going on, or what is being taught in class.
- lostdog 8y ago
  Stuyvesant is a high school, so it's in the middle of the pipeline and not at the end.
icelancer 8y ago
>> Test scores are race-blind and fair. Kind of. There is a fair criticism to be made about tests being unintentionally culturally slanted towards the people that make them, often white Americans. Of course, in this case, your point is correct, since the people that are overindexing on the tests are Asian-Americans.
- iosDrone 8y ago
  Can you provide some examples of cultural bias? People almost never provide examples. I thought that was always a bullshit excuse, especially considering that these test scores closely track IQ scores.
  - aavalle 8y ago
    While the individual questions may be (for the most part) purged of any cultural bias, family income remains the primary demographic correlated with test performance. https://www.washingtonpost.com/news/wonk/wp/2014/03/05/these-four-charts-show-how-the-sat-favors-the-rich-educated-families/?noredirect=on&utm_term=.a95b2a0523d3 https://www.washingtonpost.com/news/wonk/wp/2014/03/05/these... https://pdfs.semanticscholar.org/e7a0/dfdee8ad5b91e1d179e9e2e838a4952f00be.pdf https://pdfs.semanticscholar.org/e7a0/dfdee8ad5b91e1d179e9e2...
    - graycat 8y ago
      Of course, a correlation with income is likely a spurious correlation from more basic causes. E.g., suppose a poor family whose kids are doing poorly in school won the lottery and suddenly has high income; do we expect that suddenly then their kids will do well in school?
    - mrchicity 8y ago
      Is that surprising though? There's a strong correlation between pay and intellectual ability/education. There aren't many dull, uneducated engineers, doctors, lawyers, or executives. If intelligence is at all heritable, you'd expect to see this. If educated parents tend to push their own children to become more educated, you'd expect to see this. I wonder what the chart looks like broken down by where parents went to school or some non-financial metric. I would bet the children of well-educated college professors or journalists also score well, despite growing up with a more middle class household income.
mrchicity 8y ago
I agree totally. Tests can be slightly flawed, and rich kids can prep their way to another 50 points, but nothing else in life comes close in terms of fairness. Maybe sports and that's it. There's something noble about that. Tell me where else a poor farmer's kid from Alabama, a child of refugees who spoke no English at home, someone from an unpopular religion, a kid who grew up in the ghetto without a father, someone from the working class who speaks with a regional accent, etc. etc. can compete on a pretty level playing field with wealthy elites?
Fomite 8y ago
"Test scores are race-blind and fair." This is not axiomatically true.
- mrchicity 8y ago
  Why are they unfair? I've heard the old saw about the 1980s SATs asking some analogy with a regatta:yacht, but what bias has been uncovered lately? And why is this bias mostly helping Asians, a minority group, who more likely to be learning English as a second language?
  - iosDrone 8y ago
    I was under the impression that the test-makers have scrupulously avoided culturally-biased questions for decades, and yet left-leaning people casually continue tossing it around as part of their war on merit.
    - icelancer 8y ago
      I avoid eating too many potato chips and I still overconsume them. Just because you know you have the bias doesn't mean you defeat it 100% of the time. I agree that the effect is small if anything in today's society, however.
      - iosDrone 8y ago
        Provide some recent examples of bias on tests.
  - Fomite 8y ago
    There are a number of just-so stories you can make about them being either unfair or fair - but the education field doesn't consider this a settled question, and unless you're looking at the data in a systematic fashion with a clearly defined outcome and access to lots of control variables, you can't actually be sure the test yields the "best" students. And indeed, the definition of "best" in this context is circular. I'm not saying you're wrong, just that you can't be certain you're right.
otoburb 8y ago
>>[...] but they should encourage the underrepresented ethnicities to do better, [...] This policy change is exactly that - an attempt to encourage poorly performing students to do better, where a large majority of poorly performing students also happen to be underrepresented ethnicities. The NYC Department of Education has and continues to try many actions to increase educational outcomes for their students overall -- de Blasio is trying to accelerate the process before he's out of office.

swsieber 8y ago

This. The end of the pipeline is not the correct place to change pipeline demographics, and in cases like this, is actually harmful.

squiggleblaz 8y ago
But where is the right place to change pipeline demographics? The problem is that certain groups are intergenerationally poor. One mechanism that is available to us to fix that is education. So this isn't the end of the pipeline, it's just the start of the next pipe. Where do you propose we intervene? Should we give a cushy job to the parents of every disadvantaged minority person as soon as they're born? Maybe we should give an education to the parents of every disadvantaged minority person before they're even the apple in the eye of said parents?
- confiscate 8y ago
  You intervene when the student is still learning the material, before the test. You do not intervene after the test. At that point the learning process for that material is already over. The student has already "missed out" on the learning opportunity. You would just be giving a good grade to someone who doesn't know the material. Yes we need to help folks who are intergenerationally poor. But, you do that by helping their kids learn better and meet the testing bar. You are not helping them by lowering the bar for them. Giving the kids better grades even if they don't know the material, would actually hurt them.
  - ytpete 8y ago
    > You are not helping them by lowering the bar for them. But you are helping underperforming students by placing them into top schools, comingled with better students - lots of research shows that. So lowering the bar for admission really does help. I think the open question is how that balances against the cost of having fewer slots available at those top schools for the very best students.
    - confiscate 8y ago
      I don't know what research you are pointing to. But I am pretty sure that if, say, a student can't pass a multiplication test, but you give that student an A+ anyways, and then let that student into a "top calculus class", that's going to hurt the student more than it helps. Ya sure the student gets to mingle with better students. But what you would usually get is, the student would end up being someone who mentions a lot of buzz words like "fractal integrals" just to fit in, but have zero idea what is going on, or what is being taught in class.
lostdog 8y ago
Stuyvesant is a high school, so it's in the middle of the pipeline and not at the end.

icelancer 8y ago

>> Test scores are race-blind and fair. Kind of. There is a fair criticism to be made about tests being unintentionally culturally slanted towards the people that make them, often white Americans. Of course, in this case, your point is correct, since the people that are overindexing on the tests are Asian-Americans.

iosDrone 8y ago
Can you provide some examples of cultural bias? People almost never provide examples. I thought that was always a bullshit excuse, especially considering that these test scores closely track IQ scores.
- aavalle 8y ago
  While the individual questions may be (for the most part) purged of any cultural bias, family income remains the primary demographic correlated with test performance. https://www.washingtonpost.com/news/wonk/wp/2014/03/05/these-four-charts-show-how-the-sat-favors-the-rich-educated-families/?noredirect=on&utm_term=.a95b2a0523d3 https://www.washingtonpost.com/news/wonk/wp/2014/03/05/these... https://pdfs.semanticscholar.org/e7a0/dfdee8ad5b91e1d179e9e2e838a4952f00be.pdf https://pdfs.semanticscholar.org/e7a0/dfdee8ad5b91e1d179e9e2...
  - graycat 8y ago
    Of course, a correlation with income is likely a spurious correlation from more basic causes. E.g., suppose a poor family whose kids are doing poorly in school won the lottery and suddenly has high income; do we expect that suddenly then their kids will do well in school?
  - mrchicity 8y ago
    Is that surprising though? There's a strong correlation between pay and intellectual ability/education. There aren't many dull, uneducated engineers, doctors, lawyers, or executives. If intelligence is at all heritable, you'd expect to see this. If educated parents tend to push their own children to become more educated, you'd expect to see this. I wonder what the chart looks like broken down by where parents went to school or some non-financial metric. I would bet the children of well-educated college professors or journalists also score well, despite growing up with a more middle class household income.

mrchicity 8y ago

I agree totally. Tests can be slightly flawed, and rich kids can prep their way to another 50 points, but nothing else in life comes close in terms of fairness. Maybe sports and that's it. There's something noble about that. Tell me where else a poor farmer's kid from Alabama, a child of refugees who spoke no English at home, someone from an unpopular religion, a kid who grew up in the ghetto without a father, someone from the working class who speaks with a regional accent, etc. etc. can compete on a pretty level playing field with wealthy elites?

Fomite 8y ago

"Test scores are race-blind and fair." This is not axiomatically true.

mrchicity 8y ago
Why are they unfair? I've heard the old saw about the 1980s SATs asking some analogy with a regatta:yacht, but what bias has been uncovered lately? And why is this bias mostly helping Asians, a minority group, who more likely to be learning English as a second language?
- iosDrone 8y ago
  I was under the impression that the test-makers have scrupulously avoided culturally-biased questions for decades, and yet left-leaning people casually continue tossing it around as part of their war on merit.
  - icelancer 8y ago
    I avoid eating too many potato chips and I still overconsume them. Just because you know you have the bias doesn't mean you defeat it 100% of the time. I agree that the effect is small if anything in today's society, however.
    - iosDrone 8y ago
      Provide some recent examples of bias on tests.
- Fomite 8y ago
  There are a number of just-so stories you can make about them being either unfair or fair - but the education field doesn't consider this a settled question, and unless you're looking at the data in a systematic fashion with a clearly defined outcome and access to lots of control variables, you can't actually be sure the test yields the "best" students. And indeed, the definition of "best" in this context is circular. I'm not saying you're wrong, just that you can't be certain you're right.

otoburb 8y ago

>>[...] but they should encourage the underrepresented ethnicities to do better, [...] This policy change is exactly that - an attempt to encourage poorly performing students to do better, where a large majority of poorly performing students also happen to be underrepresented ethnicities. The NYC Department of Education has and continues to try many actions to increase educational outcomes for their students overall -- de Blasio is trying to accelerate the process before he's out of office.