Although there’s more to politics than race, race is an important part of it, and we’re obligated to treat other people fairly with respect to race. The argument that it doesn’t matter how racially biased a test is because it’s good in other ways isn’t something I am inclined to accept.
The argument that it doesn’t matter how racially biased a test is because it’s good in other ways isn’t something I am inclined to accept.
I assume this is hyperbole, since obviously a truly perfect test could draw from any subset of the population, as long as it was large enough to contain near-perfect individuals.
With that said, I agree, we should attempt to avoid any bias in such a test, including that of race (I would not, however, single this possibility out.) That is what I meant by
That said, you would definitely have to be careful to ensure the test was as good as possible.
However, beyond a certain level of conscientiousness, demanding perfectly unbiased tests becomes counterproductive; especially when one focuses on one possible bias to the exclusion of others. In truth, even age is a racially biased criterion.
In context, MugaSofer had claimed that if a test that allows young people to vote based on IQ tests black people of equal intelligence as 5 points lower IQ, that’s okay because an age test is worse than that. I was, therefore, referring to that kind of bias. I’m not sure whether you would call “gives a number 5 points lower for black people of equal intelligence” ‘how the test works’ or ‘which outcomes it produces’.
In this context, MugaSofer’s test is clearly “how it works” because the test explicitly looks at the color of skin and subtracts 5 from the score if the skin is dark enough.
On the other hand, “which outcomes it produces” is the more or less standard racial bias test applied by government agencies to all kinds of businesses and organizations.
I didn’t describe a test which looks at the color of skin and subtracts 5; I described a test which produces results 5 points lower for people with a certain color of skin. Whether it does that by looking at the color of skin explicitly, or by being an imperfect measure of intelligence where the imperfection is correlated to skin color, I didn’t specify, and I was in fact thinking of the latter case.
These are two rather different things. I am not sure how the latter case works—if the test is blinded to the skin color but you believe it discriminates against blacks, (1) How do you know the “true” IQ which the test understates; and (2) what is it, then, that the test picks up as a proxy or correlate to the skin color?
Standard IQ tests show dependency on race—generally the mean IQ of blacks is about one standard deviation below the mean IQ of whites.
In my experience, if someone is claiming that a test is racially biased, they are claiming that properly understanding the question requires cultural context which is more or less common in one race than another.
An example I found here is a multiple-choice question which asks the student to select the pair of words with a relationship similar to the relationship between a runner and a marathon. The correct answer there was “oarsman” and “regatta”. Clearly, there was a cultural context required to correctly answer this question; examining the correlations between socioeconomic status and race, I would expect to find that the cultural context is more common among rich caucasians.
In my experience, if someone is claiming that a test is racially biased, they are claiming that properly understanding the question requires cultural context which is more or less common in one race than another.
In my experience if someone is claiming that a test is racially biased, they just don’t like the test results. Not always, of course, but often enough.
is more common among rich caucasians
Then the fact that East Asian people show mean IQ noticeably higher than that of caucasians would be a bit inconvenient, wouldn’t it?
In my experience if someone is claiming that a test is racially biased, they just don’t like the test results. Not always, of course, but often enough.
and
Steelman this.
What exactly do you mean by “often enough”? Do you mean to say that there is such a large number of false positives in claims of racial bias that none of them should be investigated? I am confused by your dismissal of this phenomenon.
Regarding the fact that East Asians tend to score higher than Caucasians on IQ tests (I am familiar with this difference in the US; I do not know if it applies to comparison between East Asian and majority-Caucasian countries), I would attribute it to culture and self-selection.
In the case of the United States, it is my understanding that immigration from Europe dominated immigration to the US during the Industrial Revolution—when the US was looking for, and presumably attracting, manual laborers—while recently, immigrants from Asia have made up a far larger share of the total immigrants to the US. I would guess that relative to European-Americans*, Asian-Americans’ immigrant ancestors are more likely to have self-selected for the ability to compete in an intelligence-based trade. This selection bias, propagating through to descendants (intelligent people tend to have intelligent children), would seem to at least partially explain why Asian-Americans score higher.
I do not have any information on Caucasians in their ancestral homelands vs. East Asians in their ancestral homelands.
*Based on recollection of stories told to me and verified only by a quick check online, so if others could chime in with supporting/opposing evidence, that would be appreciated.
I mean that a large number of different studies over several decades using different methodologies in various countries came up with the same results: the average IQ of people belonging to different gene pools (some of which match the usual idea of race and some do not) is not the same.
That finding happens to be ideologically or morally unacceptable to a large number of people. Normally they just ignore it, but when when they have to confront it the typical reaction—one that happens “often enough”—is denial: the test is racially biased and so invalid. Example: you.
Do you mean to say that there is such a large number of false positives in claims of racial bias that none of them should be investigated?
I do not believe I have said anything even remotely resembling this.
I am familiar with this difference in the US; I do not know if it applies to comparison between East Asian and majority-Caucasian countries
Yes, it does apply.
I would attribute it to culture and self-selection
Before you commit to defending a position, it’s useful to do a quick check to see whether it’s defensible. You think no one ran any IQ studies in China?
Thank you for clarifying your points. I mistakenly interpreted “often enough” as indicating some threshold of frequency of false positives beyond which it would not be appropriate to take the problem seriously. I apologize for arguing a straw man.
I was considering mostly the difference among people of different races in the United States, as I assumed that would minimize the effects of cultural difference (though not eliminate it) on the intelligence of the participants and their test results. I would anticipate that cultural influences do affect a person’s intelligence—the hypothetical quality which we imperfectly measure, not the impact that quality leaves on a test—as it can motivate certain avenues of self-improvement through its values, or simply allow access to different resources.
I am not surprised that there are IQ differences among racial groups. In fact, I would be shocked to learn that every culture and every natural environment and every historical happening in the entirety of human civilization happened to produce the exact same level of average intelligence. I would be surprised, but not shocked, to learn that there existed a strong, direct causation between race (as a genetic difference rather than a social phenomenon) and intelligence.
I did not mean to imply that because a test outputs different results for different racial groups, that it must be biased. I merely meant to say that bias can exist, though I am not certain whether or not it does, or to what degree. All in all, I seem to have made rather a fool of myself, jumping at shadows, and for that I am sorry.
An example I found here is a multiple-choice question which asks the student to select the pair of words with a relationship similar to the relationship between a runner and a marathon. The correct answer there was “oarsman” and “regatta”. Clearly, there was a cultural context required to correctly answer this question; examining the correlations between socioeconomic status and race, I would expect to find that the cultural context is more common among rich caucasians.
I’ve never seen any question resembling this on any IQ test I’ve ever taken. Have you? (Note that your link refers to the SAT I, which is not an IQ test.)
Is anyone claiming that the WAIS, for instance, is culturally biased in a similar way?
Feel free to propose that in fact it doesn’t matter how racially biased a test is because it’s good in other ways. I don’t know how many people will agree with you, though.
He claimed that a test that is bad overall is worse than a racially biased test. That might be a nontrivial argument if it he could show that it is worse by some fairly universal criterion. I pointed out that that he can’t show this, because I can come up with a scenario where the racially biased test is clearly worse than the overall bad test.
His reply to that was “there is more to politics than race”. In context (rather than by taking the literal words), he’s telling me that I shouldn’t emphasize race so much when talking politics. His argument for that? Um… none, really. There’s no argument to respond to or accept. All I can do is say “no, I don’t accept that premise. I think my emphasis on race is appropriate”.
If I may jump in here … Eugene seems to be asking if you consider non-racism inherently, terminally important or purely instrumental in the great war against sucky tests.
You seem to be agreeing that yes, racism really is more important than, say, conservative bias.
I’m not certain if you actually believe that … I would guess you do … but you seemed somewhat confused by the question, so I thought I’d ask.
You realize there’s more to politics than race, right?
That said, you would definitely have to be careful to ensure the test was as good as possible.
Although there’s more to politics than race, race is an important part of it, and we’re obligated to treat other people fairly with respect to race. The argument that it doesn’t matter how racially biased a test is because it’s good in other ways isn’t something I am inclined to accept.
I assume this is hyperbole, since obviously a truly perfect test could draw from any subset of the population, as long as it was large enough to contain near-perfect individuals.
With that said, I agree, we should attempt to avoid any bias in such a test, including that of race (I would not, however, single this possibility out.) That is what I meant by
However, beyond a certain level of conscientiousness, demanding perfectly unbiased tests becomes counterproductive; especially when one focuses on one possible bias to the exclusion of others. In truth, even age is a racially biased criterion.
Do you define racial bias by how the test works or by which outcomes it produces?
In context, MugaSofer had claimed that if a test that allows young people to vote based on IQ tests black people of equal intelligence as 5 points lower IQ, that’s okay because an age test is worse than that. I was, therefore, referring to that kind of bias. I’m not sure whether you would call “gives a number 5 points lower for black people of equal intelligence” ‘how the test works’ or ‘which outcomes it produces’.
In this context, MugaSofer’s test is clearly “how it works” because the test explicitly looks at the color of skin and subtracts 5 from the score if the skin is dark enough.
On the other hand, “which outcomes it produces” is the more or less standard racial bias test applied by government agencies to all kinds of businesses and organizations.
I didn’t describe a test which looks at the color of skin and subtracts 5; I described a test which produces results 5 points lower for people with a certain color of skin. Whether it does that by looking at the color of skin explicitly, or by being an imperfect measure of intelligence where the imperfection is correlated to skin color, I didn’t specify, and I was in fact thinking of the latter case.
These are two rather different things. I am not sure how the latter case works—if the test is blinded to the skin color but you believe it discriminates against blacks, (1) How do you know the “true” IQ which the test understates; and (2) what is it, then, that the test picks up as a proxy or correlate to the skin color?
Standard IQ tests show dependency on race—generally the mean IQ of blacks is about one standard deviation below the mean IQ of whites.
In my experience, if someone is claiming that a test is racially biased, they are claiming that properly understanding the question requires cultural context which is more or less common in one race than another.
An example I found here is a multiple-choice question which asks the student to select the pair of words with a relationship similar to the relationship between a runner and a marathon. The correct answer there was “oarsman” and “regatta”. Clearly, there was a cultural context required to correctly answer this question; examining the correlations between socioeconomic status and race, I would expect to find that the cultural context is more common among rich caucasians.
In my experience if someone is claiming that a test is racially biased, they just don’t like the test results. Not always, of course, but often enough.
Then the fact that East Asian people show mean IQ noticeably higher than that of caucasians would be a bit inconvenient, wouldn’t it?
I’d like to quote you twice:
and
What exactly do you mean by “often enough”? Do you mean to say that there is such a large number of false positives in claims of racial bias that none of them should be investigated? I am confused by your dismissal of this phenomenon.
Regarding the fact that East Asians tend to score higher than Caucasians on IQ tests (I am familiar with this difference in the US; I do not know if it applies to comparison between East Asian and majority-Caucasian countries), I would attribute it to culture and self-selection.
In the case of the United States, it is my understanding that immigration from Europe dominated immigration to the US during the Industrial Revolution—when the US was looking for, and presumably attracting, manual laborers—while recently, immigrants from Asia have made up a far larger share of the total immigrants to the US. I would guess that relative to European-Americans*, Asian-Americans’ immigrant ancestors are more likely to have self-selected for the ability to compete in an intelligence-based trade. This selection bias, propagating through to descendants (intelligent people tend to have intelligent children), would seem to at least partially explain why Asian-Americans score higher.
I do not have any information on Caucasians in their ancestral homelands vs. East Asians in their ancestral homelands.
*Based on recollection of stories told to me and verified only by a quick check online, so if others could chime in with supporting/opposing evidence, that would be appreciated.
I mean that a large number of different studies over several decades using different methodologies in various countries came up with the same results: the average IQ of people belonging to different gene pools (some of which match the usual idea of race and some do not) is not the same.
That finding happens to be ideologically or morally unacceptable to a large number of people. Normally they just ignore it, but when when they have to confront it the typical reaction—one that happens “often enough”—is denial: the test is racially biased and so invalid. Example: you.
I do not believe I have said anything even remotely resembling this.
Yes, it does apply.
Before you commit to defending a position, it’s useful to do a quick check to see whether it’s defensible. You think no one ran any IQ studies in China?
Thank you for clarifying your points. I mistakenly interpreted “often enough” as indicating some threshold of frequency of false positives beyond which it would not be appropriate to take the problem seriously. I apologize for arguing a straw man.
I was considering mostly the difference among people of different races in the United States, as I assumed that would minimize the effects of cultural difference (though not eliminate it) on the intelligence of the participants and their test results. I would anticipate that cultural influences do affect a person’s intelligence—the hypothetical quality which we imperfectly measure, not the impact that quality leaves on a test—as it can motivate certain avenues of self-improvement through its values, or simply allow access to different resources.
I am not surprised that there are IQ differences among racial groups. In fact, I would be shocked to learn that every culture and every natural environment and every historical happening in the entirety of human civilization happened to produce the exact same level of average intelligence. I would be surprised, but not shocked, to learn that there existed a strong, direct causation between race (as a genetic difference rather than a social phenomenon) and intelligence.
I did not mean to imply that because a test outputs different results for different racial groups, that it must be biased. I merely meant to say that bias can exist, though I am not certain whether or not it does, or to what degree. All in all, I seem to have made rather a fool of myself, jumping at shadows, and for that I am sorry.
I’ve never seen any question resembling this on any IQ test I’ve ever taken. Have you? (Note that your link refers to the SAT I, which is not an IQ test.)
Is anyone claiming that the WAIS, for instance, is culturally biased in a similar way?
What’s your counter-argument?
It’s not an argument, it’s a premise.
Feel free to propose that in fact it doesn’t matter how racially biased a test is because it’s good in other ways. I don’t know how many people will agree with you, though.
You said you weren’t willing to accept the argument. Do you have any better reason than “I don’t feel like it”?
Wasn’t willing to accept what argument?
He claimed that a test that is bad overall is worse than a racially biased test. That might be a nontrivial argument if it he could show that it is worse by some fairly universal criterion. I pointed out that that he can’t show this, because I can come up with a scenario where the racially biased test is clearly worse than the overall bad test.
His reply to that was “there is more to politics than race”. In context (rather than by taking the literal words), he’s telling me that I shouldn’t emphasize race so much when talking politics. His argument for that? Um… none, really. There’s no argument to respond to or accept. All I can do is say “no, I don’t accept that premise. I think my emphasis on race is appropriate”.
Why is bias on the test that happens to correlate with race worse than any other bias?
I don’t see any argument in that.
If I may jump in here … Eugene seems to be asking if you consider non-racism inherently, terminally important or purely instrumental in the great war against sucky tests.
You seem to be agreeing that yes, racism really is more important than, say, conservative bias.
I’m not certain if you actually believe that … I would guess you do … but you seemed somewhat confused by the question, so I thought I’d ask.