"And, are IQ test biased for and against certain cultures and backgrounds? Absolutely. It's not intentional, but it's unavoidable."
You provide no evidence for this claim. To claim that a test is "biased" against black testees means that it is not a valid predictor of black testees' performance. This is not true. SAT's, ACT's, Miller Analogy Tests, and other ability tests are strong predictors of academic performance, especially when coupled with high school GPA's, and this is why colleges have used them.
If SAT's plus GPA's were biased towards black high school graduates seeking admission to colleges, we would expect to see more black students who scored low on the tests going on to succeed in college (thereby exceeding expectations set by their test scores and their high school GPA's). We do not see this happening. High school students of any color who score low on college admission tests and have low GPA's are less likely to succeed in college and more likely to drop out before completing their degrees than are students with higher high school GPA's and higher test scores.
The combination of these numbers does not explain ALL of the variation in who succeeds in college and who does not, but the predictive validity of the tests plus GPA is high. This is a matter of fact, not of opinion. It happens to be a fact that many people would prefer to deny, because it might mean that black students are innately inferior intellectually to whites, Chinese, and so on, and that white students are inferior to Chinese and so on. Reality is not always what everyone wants it to be.
It is true that "innate general ability" is the thing that we are trying to measure with IQ tests, and we still need another measure to define what we mean by ability. IQ scores at the low end of the range reflect peoples' abilities to function independently and to benefit from training in basic life skills, for example. At the higher end of the range, which is what Loury is generally talking about, IQ tests and college admission tests combined with measures of past achievement predict performance in academia and knowledge-based careers. The best predictors of performance in jobs are tests that measure abilities that will be needed in those jobs, so SAT's are not generally used to select people interested in skilled trades. Another example would be Art or Music careers, for which auditions and portfolios are used to predict performance.
It is wonderful that you are open to various viewpoints, but with respect to the predictive validity versus invalidity or "bias" of tests, this can be determined by observable outcomes. People who have low SAT's and low high school GPA's have much lower odds of succeeding in college than people who have high ones, and this is a fact.
Excellent explanation of the predictive validity in academic performance associated with IQ and GPA. You managed to make a complex and confusing relationship crystal clear—nice work!
"And, are IQ test biased for and against certain cultures and backgrounds? Absolutely. It's not intentional, but it's unavoidable."
You provide no evidence for this claim. To claim that a test is "biased" against black testees means that it is not a valid predictor of black testees' performance. This is not true. SAT's, ACT's, Miller Analogy Tests, and other ability tests are strong predictors of academic performance, especially when coupled with high school GPA's, and this is why colleges have used them.
If SAT's plus GPA's were biased towards black high school graduates seeking admission to colleges, we would expect to see more black students who scored low on the tests going on to succeed in college (thereby exceeding expectations set by their test scores and their high school GPA's). We do not see this happening. High school students of any color who score low on college admission tests and have low GPA's are less likely to succeed in college and more likely to drop out before completing their degrees than are students with higher high school GPA's and higher test scores.
The combination of these numbers does not explain ALL of the variation in who succeeds in college and who does not, but the predictive validity of the tests plus GPA is high. This is a matter of fact, not of opinion. It happens to be a fact that many people would prefer to deny, because it might mean that black students are innately inferior intellectually to whites, Chinese, and so on, and that white students are inferior to Chinese and so on. Reality is not always what everyone wants it to be.
It is true that "innate general ability" is the thing that we are trying to measure with IQ tests, and we still need another measure to define what we mean by ability. IQ scores at the low end of the range reflect peoples' abilities to function independently and to benefit from training in basic life skills, for example. At the higher end of the range, which is what Loury is generally talking about, IQ tests and college admission tests combined with measures of past achievement predict performance in academia and knowledge-based careers. The best predictors of performance in jobs are tests that measure abilities that will be needed in those jobs, so SAT's are not generally used to select people interested in skilled trades. Another example would be Art or Music careers, for which auditions and portfolios are used to predict performance.
It is wonderful that you are open to various viewpoints, but with respect to the predictive validity versus invalidity or "bias" of tests, this can be determined by observable outcomes. People who have low SAT's and low high school GPA's have much lower odds of succeeding in college than people who have high ones, and this is a fact.
https://www.newsweek.com/when-race-trumps-merit-how-pursuit-equity-threatens-lives-opinion-1794872. Saw this today on realclearpolitics and looks related to the discussion on tests.
Yep! When I was a college student, there was a slogan "Don't trust anyone over 30." I am thinking, "Don't trust any doctor under 50."
Thanks for forwarding the article--Heather MacDonald's articles are always worth reading.
Excellent explanation of the predictive validity in academic performance associated with IQ and GPA. You managed to make a complex and confusing relationship crystal clear—nice work!
Thanks, Richard!