> 1. Taking a test on multiple days produces high enough variance in results, that the test results are not predictive of any sort of ability or future outcomes
What I'm saying here is that hunger and sleep are two confounding factors always present in any testing situation, at least in my experience as an educator. I have seen it personally in my students, especially when comparing my 8am section performance with my 12pm section's performance for the same class on the same tests and material. The 8am section consistently performs worse (sometimes by half a letter grade or more) on their tests compared to their 12pm peers, and this has been true across my career. What's interesting is I can compare these students across semesters, since they take multiple classes with me. I can see how they perform at 8am one semester vs 12pm another. But this is just my own little experience with college-age students, I'm not familiar with recent literature.
Anyway, the point is if I'm seeing these things in my students there is no reason to thing we wouldn't see these performance drops on 3rd graders doing an IQ test as well. So the question is: why aren't they controlling for these things? Or are they, and I don't know. Because they didn't control them for my cohort when I was tested.
> 2. The majority of 'children of color' are hungry when taking IQ tests ('of color' in popular usage means non-white) 3. The majority of white children are well-fed when taking IQ tests
I wasn't making a generalization here, I was specifically talking about my school.
> Reading between the lines, it seems you are hinting at the idea that all humans who approach the IQ test are equal in ability that the IQ test is testing for. If there are any disparities in results amongst any groups, it must be the result of other factors like being hungry/tired/etc. Is that a fair characterization?
Don't read between the lines here too much, I was making a concrete argument: If we want to measure something like IQ, we should do so rigorously and with purpose, because otherwise what are we even doing?
Even if we just assume IQ is defined as the result of the IQ test, we know that test result can vary based on testing conditions and the emotional state of the tester during the test. With regular testing this is not so bad because the effects average out over time, unless the problem is chronic. With IQ test though, the presence of these confounding factors is especially vexatious, because there is a general perception that IQ is a fixed, immutable quantity, so the score you get on the test cannot be improved.
In the context of gifted programs it means these tests are usually administered once and early during the normal course of a child's 12 year education, and these program really only exist in elementary school; by the time children get to middle and high school there are other mechanisms for sorting children (standard, honors, ap, electives, votech etc.). With something so consequential, we need to do it right, rather than just do a thing because it sounds right and hope that it works out for the best for everyone. If the consequences weren't so monumental than the slapdash approach I experienced (and again, this was a long time ago but at the same time not so long ago. Maybe it's different now, but this is how it was for me maybe things have changed drastically).
Despite the drawbacks you listed here of the IQ test (which are drawbacks of testing in general I think), which are all true, there is no better predictor we know of than SAT like tests or IQ (which are pretty similar and what they test for). What do you suggest doing then?
What I'm saying here is that hunger and sleep are two confounding factors always present in any testing situation, at least in my experience as an educator. I have seen it personally in my students, especially when comparing my 8am section performance with my 12pm section's performance for the same class on the same tests and material. The 8am section consistently performs worse (sometimes by half a letter grade or more) on their tests compared to their 12pm peers, and this has been true across my career. What's interesting is I can compare these students across semesters, since they take multiple classes with me. I can see how they perform at 8am one semester vs 12pm another. But this is just my own little experience with college-age students, I'm not familiar with recent literature.
Anyway, the point is if I'm seeing these things in my students there is no reason to thing we wouldn't see these performance drops on 3rd graders doing an IQ test as well. So the question is: why aren't they controlling for these things? Or are they, and I don't know. Because they didn't control them for my cohort when I was tested.
> 2. The majority of 'children of color' are hungry when taking IQ tests ('of color' in popular usage means non-white) 3. The majority of white children are well-fed when taking IQ tests
I wasn't making a generalization here, I was specifically talking about my school.
> Reading between the lines, it seems you are hinting at the idea that all humans who approach the IQ test are equal in ability that the IQ test is testing for. If there are any disparities in results amongst any groups, it must be the result of other factors like being hungry/tired/etc. Is that a fair characterization?
Don't read between the lines here too much, I was making a concrete argument: If we want to measure something like IQ, we should do so rigorously and with purpose, because otherwise what are we even doing?
Even if we just assume IQ is defined as the result of the IQ test, we know that test result can vary based on testing conditions and the emotional state of the tester during the test. With regular testing this is not so bad because the effects average out over time, unless the problem is chronic. With IQ test though, the presence of these confounding factors is especially vexatious, because there is a general perception that IQ is a fixed, immutable quantity, so the score you get on the test cannot be improved.
In the context of gifted programs it means these tests are usually administered once and early during the normal course of a child's 12 year education, and these program really only exist in elementary school; by the time children get to middle and high school there are other mechanisms for sorting children (standard, honors, ap, electives, votech etc.). With something so consequential, we need to do it right, rather than just do a thing because it sounds right and hope that it works out for the best for everyone. If the consequences weren't so monumental than the slapdash approach I experienced (and again, this was a long time ago but at the same time not so long ago. Maybe it's different now, but this is how it was for me maybe things have changed drastically).