> Let's put this in the proper context. Many modern studies and analysis show similar disparities in intellectual abilities between ethnicities and races, based on a number of tests designed in the modern day by many different institutions globally. The suggestion that IQ tests are on the whole are a racist ploy is very conspiratorial.
The parent comment that original suggested this claim linked to an article tracing the academic influences of the I.Q. test and other similar standardized testing frameworks, and it doesn't seem "very conspiratorial" to me in the absence of any specific critiques.
> Furthermore, it's pretty hard to imagine how one would design visuospatial and pattern matching problems so that they are discriminatory.
Maybe my imagination is more active than yours, but it doesn't seem that hard to imagine for me. Broadly speaking, I'd imagine that the process of coming up with any sort of test like this would consist of the following steps:
1. Come up with potential questions for the test
2. Assess people with the test in a controlled environment
3. Revise the set of questions based on the results of step 2, then repeat steps 2 and 3 until you're satisfied
To tailor the results to a certain group, give the test to both people in and outside of that group in step 2, then in step 3 revise the questions towards the type that the favored group scored better on (e.g. throw out questions that the disfavored group scored better on, come up with more questions similar to the type that the favored group scored better on, etc.).
You wouldn't be guaranteed to find a set of questions that ended up statistically favorable to the favored group, but it certainly doesn't seem very hard to imagine it might work sometimes, and for it to have happened, it only would need to have worked once.
there is no “the I.Q. test” there are many different ones, designed differently, the article distorts this, and attacks the original Stanford-Binet test as if it was the universe of IQ testing, and even there the entire substance of its attack on that test, delivered in meandering back and forth where it drifts in and out of discussing the SAT and other tests, seems to be that a researcher who later was influential in developing other tests also wrote their dissertation about the same work by Alfred Binet that a completely different researcher leveraged in developing what became the Stanford-Binet IQ test was a eugenicist who thought that blacks and other races were inferior and that testing supported this. Which isn’t just ad hominem, its ad hominem directed at someone not even involved in the development of the test. There is literally nothing in the piece actually arguing for any ill-motive, much less actual scientific flaw, in IQ testing.
There are some real issues that the article hints at – particularly regarding school rankings – but they spent a lot of time trying to juxtapose negative things with various tests that they had no real argument about and just hoped to use guilt-by-(often strained)-association to impugn.
The parent comment that original suggested this claim linked to an article tracing the academic influences of the I.Q. test and other similar standardized testing frameworks, and it doesn't seem "very conspiratorial" to me in the absence of any specific critiques.
> Furthermore, it's pretty hard to imagine how one would design visuospatial and pattern matching problems so that they are discriminatory.
Maybe my imagination is more active than yours, but it doesn't seem that hard to imagine for me. Broadly speaking, I'd imagine that the process of coming up with any sort of test like this would consist of the following steps:
To tailor the results to a certain group, give the test to both people in and outside of that group in step 2, then in step 3 revise the questions towards the type that the favored group scored better on (e.g. throw out questions that the disfavored group scored better on, come up with more questions similar to the type that the favored group scored better on, etc.).You wouldn't be guaranteed to find a set of questions that ended up statistically favorable to the favored group, but it certainly doesn't seem very hard to imagine it might work sometimes, and for it to have happened, it only would need to have worked once.