In addition to what ucbalumnus noted above, the SAT/ACT attempts to measure what a student has learned. That brings in the whole advantaged/disadvantaged (in terms of k-12 education) concept. What we know for sure, from data, is that ACT/SAT scores are correlated with family income.
1 Like
Family income has a pretty high correlation with high scholastic aptitude.
There are a variety of plausible answers, but I would say the SAT tests a combination of:
(1) Physical and mental health on the day of the test;
(2) Test anxiety;
(3) Natural pace of work;
(4) Familiarity with the test format and various best practices for efficiently taking the test;
(5) How many teen and adult level books you have read;
(6) Mastery of the math concepts and practiced skills in the main sequence through Pre-Calc; and
(7) Mastery of foundational English stuff usually taught in Grade 9 in the standard college-prep sequence.
Then a whole ton of other stuff can contribute to various of those factors.
1 Like
Most of the top-scoring students these days on the SAT are second or third generation Americans, often of Asian descent, whose parents have worked and studied extremely hard to achieve an upper-middle-class income as engineers, doctors, etc., and who are pushing their children to succeed via similar paths. So sure, there’s some correlation with income, but causally it’s mainly about culture and focus.
1 Like
I know that the SAT moved away from testing “aptitude” when they dropped questions like quantitative reasoning (choose A or B if bigger, C if equal, D if can’t be determined) and analogies and antonyms on verbal. I also know that there was a lot of buzz around whether those types of questions were “biased”.
I have extremely mixed feelings. Because now we have a test that is “beatable” by sheer practice and by wealth. Wealth is converted into tutors, time, and practice opportunities.
Also, I am in STEM so I will speak to the math section. It doesn’t really test high school math very effectively. There’s a lot of focus on things like circle geometry relative to almost none on precalculus (and none above precalculus).
Furthermore, now calculators are allowed, which again marks privilege if kids practice with their same calculators. (I know there is also an embedded calculator and free online practice - provided a student knows this is out there and has technology, internet, and time to futz around and practice.) Meanwhile, someone decided that brain/paper skills instead of calculators are no longer valuable as a mark of aptitude, and I’m not sure I agree.
I know less about the verbal side, but vocabulary range is certainly an indicator of learning and reading, and it is really not tested in a meaningful way. The “writing” (editing) section is annoying/confusing even to kids who score close to 800.
Overall, I think the test is still a measure of wealth (if not more so) but now it is much LESS a measure of anything useful for colleges to see. Also, the score inflation means that scores are much less spread out for “top” students and therefore much less meaningful.
In the olden olden days, my parents’ scores (1950s and early 60s) had three significant figures e.g. 515. In my own day (1980s), my 1400 overall was considered excellent (and got me into M, P, and other places). There is now much less information and much less utility to the test.
I suppose this is why some schools are test optional, but that opens whole other problems. Apparently this is not a new problem. The SAT was pushed aside in favor of “holistic” admission to make sure not too many Jews were admitted almost 100 years ago… https://www.amazon.com/Chosen-History-Admission-Exclusion-Princeton/dp/0618574581
ETA: the SAT and this process in general are marked by a kind of split personality where we in the USA like to pretend that we don’t sort by ability, and that anyone can do anything with sheer grit. This is of course nonsense, as grades/scores are absolutely sorting tools, and kids start from widely variable starting lines, some of which make insurmountable differences. This is also deeply connected to (in my view poor) reforms such as 50% minimum for grades even when no work is done. All that does is narrow the band of information about a student’s achievement and inflate final grades. Much like the new SAT.
2 Likes
They have the time to focus on their kids because they don’t have to worry about working 2 jobs to make sure they have food. Families that do not have to worry about basic necessities can devote more time to actually parenting their kids. My father is an immigrant from another “work hard” culture. His parents were working all the time, and didn’t have the time to ask him if he did his homework, studied for his tests, etc.
Let’s move on from the discussion of what the SAT measures and the related correlation/causation
2 Likes
My thoughts too. Since the scattergram does not show WL it is hard to tell but the first scattergram looks like yield protection. There are three specific schools among t20 privates that have similar scattergrams (at our HS)but only one of them has the green acceptances with this low-for -T20 SAT range, the others yield protect a much smaller group at the very top.
1 Like
The parents in these cases were selected by immigration, since the immigration categories that they came through favor high educational attainment. High educational attainment of parents is probably the strongest correlate to the kids’ educational achievement. To the extent it is about culture, it is the culture of valuing education by those with high educational attainment, not the general culture of ethnic origin.
5 Likes
Family income has a pretty high correlation with the ability to provide individual attention, both in and out of the classroom, that can lead to higher performance when compared to peers who did not have those advantages. Aptitude is not determined by one’s income level.
As for the scatterplot, admissions is an art rather than a science at very selective schools. I worked in admissions at a very selective school, although it was not the type of school most on CC think of when talking about selectivity. Putting together a class with a variety of perspectives and experiences does not necessarily lend itself to allowing others to “know” who will be selected. It’s not intended to drive applicants nuts, even though it does … it’s intended to cultivate a particular group that will learn from one another as much as they learn from their professors.
3 Likes
I didn’t say that scholastic aptitude is determined by family income, of course it is not.
With respect to (some) selective schools using non-academic criteria to cultivate a group…I am curious how this is done in the top universities outside the US. In places like China, Japan, or Russia for example, do they use criteria other than tests and academics. Maybe they have tests that measure aptitude better than ours.
Scattergrams are pretty useful for the schools where a lower-bound for qualified is drawn, and everyone who meets that is admitted. It is trickier to learn much from the scattergrams for highly rejective schools, other than there is a lot more than just scores and grades that go into each decision. For some of the most selective schools there are few enough accepts from my son’s school that it was pretty easy to figure out who was who, and everyone (even those with really high grades and scores) who was accepted at schools like HYP had something extra. Here are some of the extras for high level admits that we were aware of: national level awards, URM, 1st generation citizen, legacy, recruited athlete, unique personal history that tied into intended major, and high-level ECs that tied into intended major. Many had more than one of these. In addition, test optional complicates things, as well as differences in admission requirements to certain schools or majors. I also know of a student who had a really good resume of grades, scores and activities, but submitted their application to their top choice late. I am not sure this led to their rejection directly, but it could not have helped. Every little dot on that graph has their own story.
1 Like
Actually, their tests are more heavily focused on achievement on a wide range of high school subjects. For example, the PRC gaokao testing has Chinese, math, a foreign language of the student’s choice, and then three elective subjects chosen from subjects like biology, chemistry, physics, history, geography, political science.
1 Like
Some such colleges put out that information on their web sites, so no scattergram is needed. Where the scattergram can help for these kinds of colleges is when those colleges do not publish their thresholds.
However, when such colleges admit by major, the “borderline” in the scattergram can be fuzzy, rather than a clear distinct line.
URM by itself is no longer allowed as a consideration, and 1st generation citizen is not rare among college bound students, including those aiming for the most selective universities. The other features listed certainly can have a significant effect if the college wants those things.
Some schools want applicants to have evidence of leadership, some want applicants to have done really high level (for a high school) research/internships or have similar experience, some want students from Alaska as well as Mississippi and Vermont, some want applicants with a wide variety of backgrounds and experiences, some want applicants who demonstrate a real interest in their school, some want applicants who are honed in on certain programs. What they want year to year can change. Too many CS students already enrolled may mean that they take fewer CS students in a given year. Or too many suburban students enrolled may mean a push for rural or urban students in a given year. It’s a dance, and it really does change from year to year.
2 Likes
Yes, and my husband’s acceptance was based on his French bacc exam and mine was a series of admissions tests in four different subjects.
This “building the class” and “diversity is education” is not something I have encountered in my Eastern European upbringing. If you applied to be a math major, they wanted kids who scored tops in a range of subjects pertaining to your major. It was assumed that merit was measured by subject knowledge and subject knowledge is measured by subject test. And it was assumed that you wanted to accept the academically strongest students.
I also want to point out that most kids lived at home and went to a university in the same city. College campus/dorm life was not the norm. Universities were strictly for learning.
The four years of idyllic existence on a college campus is not a universal take on education across the world for sure.
I really like A level exams in Britain. I think there is a space for more rigorous/deeper AP exams in fewer subjects (similar to A level) to be a requirement for admission.
I can see math/physics/chem/English be a requirement for STEM and a different combinations for other disciplines.
Also sports is completely divorced from academia. And musicians and artists went to their specialized universities.
scatter plot probably captures he diversity of admits and why it doesn’t correlate strictly with numerical values…
2 Likes
The people I was thinking about were admitted prior to the new rules about not considering race in admission. Regardless, URM, 1st generation citizen (with parents who were not fluent in English), and legacy by themselves did not appear to be enough for students with strong grades and scores from my son’s HS to be admitted to highly rejective schools. They still needed to have something else that was compelling.
Nor in the US, really. There are of course many US colleges where a majority, most, or indeed all of the students commute rather than living on or within walking distance of campus.
And then even among “residential” four-year colleges, there are many, many direct-admit undergrad programs (majors, specialty schools, and so on) much more focused on your academic credentials for that program, in a way that would mostly feel much more familiar to non-US families. Their scatterplots, if you will, tend to look much more like what you would expect, at least with good enough data.
But for sure some of the most Internationally-famous US colleges are firmly in what is usually called the Liberal Arts Tradition, in an exploratory form and as applied to a “residential” college. And the “top” undergrad programs in most of the rest of the world are not at all like that.
So if those are the particular US colleges you are interested in, but you are coming out of a different sort of system, that is a culture clash.
And for some people, that is part of the appeal, the opportunity for a very different sort of undergrad experience than they would get in most of the world. And conversely if you don’t actually value that, then why specifically those US colleges? At a minimum, that is a question those colleges will typically be asking–why us if you don’t really value what we do?
2 Likes
Actually, the scatterplots of colleges that admit by major or division may be less likely to show a neat division between admit and reject, even if they are purely or mainly stats-based. For example, a scatterplot of San Jose State University’ 2024 results may show a bunch of green dots down to 2.6 GPA, but also red dots up to around 4.2 GPA (let’s assume that the GPA used is as recalculated for SJSU with limited weighting with a maximum of 4.3-4.4 – note also the SJSU does not use SAT/ACT scores for admission). That may look non-sensical, but can be explained if the 4.2 rejects were applying for CS, while the 2.6 admits were applying to majors like biology, history, philosophy, and physics, which Freshmen Impaction Results | Admissions shows (divide the threshold scores by 800 to get the corresponding GPA).
Yes, part of what I meant by good enough data is if they admit by major or specialty school, you would need just the data for that major or specialty school.
I probably should not have phrased it that way because it probably sounded like I meant that might actually be common. Generally it is quite hard to get such data, so I was really just trying to make a point about how the admissions to such programs in substance would much more resemble how uni admissions work in most of the world.