I mean, not really, if you understand the issues beyond the surface level.
Usually when college admissions people say something like that about standardized tests, what they are implicitly comparing them to is something like raw HS GPAs. Raw HS GPAs can in fact be weak predictors because of the lack of standardization. So, many selective colleges have to do a lot of work to do what I would call “normalizing” transcripts, and the inputs into that process usually include the transcript, school reports, possibly counselor letter, possibly regional AO experience, possibly internal tracking data, possibly in fact test scores . . . .
So even if normalized transcripts are an important part of a college’s internal academic rating, that isn’t really a “single” indicator, and for that matter might not even be entirely separated from test scores.
Meanwhile, I am not aware of anyone seriously arguing that there is no subject matter understanding required to do well on the SAT/ACT. Indeed, they would not be learnable tests if that was not true, and they clearly are learnable to some extent.
The nuanced version of the common criticism is that at least for some people, they are also time-pressured tests. And time-pressure tests then test for rate of work and not just subject matter understanding. But they can do both.
However, the SAT/ACT need to scale to certain percentiles. And yet–at least prior to becoming more adaptive–they were limited in how far they could go in terms of subject matter understanding. So, at the highest score levels, rate of work starts to dominate who gets the highest scores among people who have all pretty much mastered the, for them, rather low-level subject matter understanding required.
But still, if you don’t get a high score, that could be for one of two reasons–you could have the subject matter understanding but a slow rate of work, or you could have a fast-enough rate of work but lack some subject matter understanding.
OK, so Yale is looking at all this and saying it wants high scores to help confirm sufficient subject matter ability, particularly in Math it seems, although that could have been just an example.
That doesn’t mean they are rejecting the observation that some people might not get high scores despite adequate subject matter ability due to slow rate of work. Instead, it is consistent with Yale not always being able to tell the difference.
So with things like this, I think you have to be looking at conditional probabilities. I am sure what you wrote is true of test-takers generally. I am not at all sure it is true of test-takers with, say, near-perfect grades in rigorous college-prep classes.
But again, it doesn’t have to be “most” to explain Yale’s preference.
Like, suppose (totally hypothetically) Yale was finding that among people who met their grade/transcript standards, 75% with low test scores also did well in their classes, and then 85% of such people with high test scores also did well in their classes.
Assuming for simplicity that there were no other factors, Yale would rationally fill up their entire admit class with people with both high grades/rigor and high test scores if they could. Because 85% is better than 75%, so why not?
Now I described this as a hypothetical, but I am actually very confident something like this is going on. And the reason I am confident about this is that after looking at a bunch of SCOIR data for our HS, it became clear to me that it was really only a small handful of colleges that seemed to have strict high test minimums for our unhooked applicants. I actually don’t have Yale data, but you could see this in the Harvard, Stanford, and MIT data, so OK, I think that is good confirmation.
But as soon as I started looking at other Ivies and Ivy+ colleges, more people were getting admitted with very high grades and somewhat lower test scores (which may or may not have actually been submitted–we do not appear to be tracking this). Even just a slight step below those few Ivy+ colleges, and many more people without high test scores were being admitted with high grades.
So what this implies to me is the following. A few colleges like Yale can fill up on just mostly the hypothetical 85% folks (meaning both really high grades and really high test scores), but then literally those people start running out. Which we know–there are only so many high test scores at all, and then only so many of them will also have really high grades, and so only so many colleges can actually fill up with such applicants.
And then other colleges have no choice but to start hedging their bets. So some kids with not the highest grades/rigor but high test scores. Some kids with not the highest test scores but high grades/rigor. In other words, they have to start admitting more and more of those hypothetical 75% kids, because there just aren’t enough of the 85% kids to go around.
So, OK, Yale and a couple of its peers can insist on having it all. But very quickly, colleges are going to have to start making some choices, and I can see in the data they are sometimes choosing high grades/rigor but maybe not the highest test scores (if they even see the test scores).
But also sometimes the opposite. So, if you are a high grades/rigor kid who can also get a really high test score, that is good. It gives you a much better shot at Yale (I think), and at least a somewhat better shot at the next couple of levels of selectivity too.
But, if you can’t get a really high test score–your odds at Yale may be very slight (unhooked), but I think you still could have a fighting chance in those next couple levels. And at a certain (still very selective) point you are going to be very likely for admittance and it won’t really matter at all.