University of California Will No Longer Consider SAT and ACT Scores
nytimes.com
nytimes.com
In terms of assessment I think grades and essays are sufficient. You might be right that it's more gameable, but is it THAT much more gameable? Should we instead waste literally tens of millions of hours of people's time so that we can say that maybe we prevented some undeserving people from entering the system? If they are truly undeserving they will fail out. And if they do well, then I guess they weren't undeserving.
I also think it's important to point out that in California, if you rank in the top 9% of your class you are guaranteed acceptance to a UC. It may not be your first choice, but if you don't get into your top choice, you can go to another UC and transfer, or go to a community college and transfer.
Exactly the same thing can be said about the college itself.
All stuff you use both in college and in the real world.
Seems a little disingenuous to discount that.
-When I took the SAT, analogies made up about one quarter of the questions on the Verbal section of the SAT. I have literally never used this skill in my life. Yet wasted an enormous amount of time memorizing vocabulary and practicing this style of question.
-The essay portion does not assess writing quality. It's literally write as fast as you can, as much as you can, and mention all the right things in a way for the grader to easily see that you have mentioned them. This is not quality writing. It's a race and knowing what to say.
-The math section is the only section that might be a little valuable. But it only tests up to geometry and little bit of Algebra 2. Most students competitively applying to UC's have taken more math than that, so their course history and grades are a better indicator of their preparedness.
So what information is the SAT actually providing to universities? In my opinion, little to nothing.
The theory was never that you literally do that contrived exercise, but that the exercise is a psychometrically valid tests of language skills that you do use.
There may be legitimate challenges to that, but “I never did that exercise outside of the context of the test” isn’t even approximately one of them.
If the SAT doesn't test actual skills NOR is a good indicator for college preparedness and success. What is it actually good for?
It seems to mainly make life easier for college admissions staff. At an enormous cost! Kids have to waste a bunch of time upping their "psychometric ability score". Parents should spend a bunch of time and money helping their kids improve their "psychometric ability score". Teachers should spend class time teaching to the "psychometric ability score" instead of actual material, because that's what gets kids to the next step.
That’s simply false. SAT scores are one of the best predictors of college preparedness and success.
The research on SAT scores and college success show very weak correlations with SAT/ACT scores and college preparedness and success. On the positive side it seems to correlate with grades in the first year of college (1). On the negative side there was no correlation between ACT scores and college graduation (2). High school GPAs, on the other hand had a strong correlation with college success.
If a metric isn't very effective at measuring what it's supposed, why collectively invest billions of dollars and hours in that metric? Just because I had to waste my time with the SAT doesn't mean we have to force that onto kids today.
1) https://www.pbs.org/wgbh/pages/frontline/shows/sats/test/wha.... 2) https://news.uchicago.edu/story/test-scores-dont-stack-gpas-...
> Results show that:
> SAT scores are strongly predictive of college performance—students with higher SAT scores are more likely to have higher grades in college. SAT scores are predictive of student retention to their second year—students with higher SAT scores are more likely to return for their sophomore year.
> SAT scores and HSGPA are both related to academic performance in college but tend to measure slightly different aspects of academic preparation. Using SAT scores in conjunction with HSGPA is the most powerful way to predict future academic performance.
> On average, SAT scores add 15% more predictive power above grades alone for understanding how students will perform in college.
> SAT scores help to further differentiate student performance in college within narrow HSGPA ranges.
In contrast, you 1) Declare anything that disagrees with you as "simply false" despite the evidence, 2) Lash out and declare me a "science denialist", and 3) To prove your point you link to the College Board, which is the company that MAKES MONEY off of the SAT test. That's like telling someone that Crest toothpaste is the best, because the Crest website said they did a study showing their toothpaste is the best, and that is science.
No, there is very little. As I said, put "sat predictive validity" in Google Scholar and see for yourself. Overwhelming consensus in the field is that the SAT is one of the best predictors of college success, alongside GPA and IQ. You cited something, but it doesn't change the scientific consensus: you can find papers published in 2021 arguing that humans have nothing to do with global warming, but the consensus is nevertheless clear. I can give you 10 citations for every one you provide, each of which will be a better designed study than you give, but you could also just put "sat predictive validity" in Google Scholar to see them.
> 2) Lash out and declare me a "science denialist"
That's because you're saying blatant untruths about the state of the science. As I said, there have been literally thousands of papers on this topic, and the conclusion is crystal clear.
> 3) To prove your point you link to the College Board, which is the company that MAKES MONEY off of the SAT test.
College Board is not a company, but in fact a non-profit. I linked it, because it is a good summary of the state of the research. I also told you where to find the actual studies, but you'll find that it is the College Board that's making the biggest and best studies on this topic, as they're obviously very interested in showing their validity. Quality of their research has, as far as I can tell, never been seriously questioned by anyone in the field. If you have any other criticism, feel free to show it, preferably in a form of paper citing College Board studies and showing the scientific misconduct you allege them to perform.
2) Nonprofits ARE companies and many of them (including the College Board) make a lot of money. 1) The College Board's total revenue is about 1 Billion dollars per year. 2) The executives earn roughly $500k - $1 million annually in total compensation (salary + bonuses). Of course, they are going to say that their product is the best. Just like Harvard, Stanford, and any educational institution (all nonprofits) are going to advertise that their methods are the best. Even the NHL is a nonprofit. Do you think the NHL is going to be the best source for research on hockey player injuries? No. Because they have a financial interest in not doing so.
https://www.washingtonpost.com/news/answer-sheet/wp/2015/09/...
The study you cite does not support that claim. See https://dynomight.net/are-tests-irrelevant/ for a close examination of it.
I can similarly counter with a study that criticizes studies that show a positive correlation with SAT and freshman GPA. For example, [1] cited 440 times performs an analysis showing that these positive correlation studies overstate their claims and over-attribute the freshman grade point average to SAT score. I am certain you can counter with another study (google scholar has 300,000+ links on the topic).
The point I am making is that the value added of SAT/ACT is unclear and not worth the time and money students spend on it. Instead, they could use SAT II or AP test scores, which are subject tests on useful material, that is already taught in schools. At least then students will spend time studying topics they may actually use in the future (chemistry, biology, US history, calculus, physics, Spanish, etc.). It would show that students can learn material on a particular topic and pass a standardized exam on it (which is most of what college is anyway).
I think a large part of why the SAT I is so controversial is because it is testing "broad abilities" and "capacity for future learning" which are definitions that changes over time and it is unclear what that actually means and how useful it is. In contrast, a subject test is clear: Can you learn material on subject X well enough to pass a test on it? If you can, well that's what completing college is like. There are studies that support the use of subject tests (SAT II) over the use of broad psychometric tests (SAT I) [2]. I think that sticking with these "broad intelligence tests" is based more on "well we've done it this way for a long time" than actual utility.
[1] https://www.sciencedirect.com/science/article/abs/pii/S03044... [2] https://www.tandfonline.com/doi/abs/10.1207/S15326977EA0801_...
I feel like the same could be argued about _all_ methods that are used. Hence the combination of various methods like SATs, grades, letters of recommendation, essays, etc. Why exactly are the SATs being singled out?
The SATs try to broadly test psychometric abilities. You have to do silly tasks like pick the best analogy. So you spend time practicing tasks that measure no particular mastery of anything. Solving crossword puzzles might be a better psychometric measure (it would also test vocabulary and it's more fun, but it would make the exam look stupid).
The math portion only tests up to Algebra 2. Many students have taken math beyond Algebra 2, so it doesn't even show colleges what their actual math ability is.
I think grades are more relevant because they show how students performed across multiple subjects over a longer period of time. Grades are more subjective and possibly gameable. But I don't think ability to answer psychometric questions is a better metric than grades.
But subject tests are not actually required right? (I honestly don't know. I just looked it up and it appeared they were not.) Should they be required going forward?
> The SATs try to broadly test psychometric abilities. You have to do silly tasks like pick the best analogy. So you spend time practicing tasks that measure no particular mastery of anything. Solving crossword puzzles might be a better psychometric measure (it would also test vocabulary and it's more fun, but it would make the exam look stupid).
Do you know that the tests don't really correlate to future success or are you speculating here? I don't take issue with your claim that e.g. analogies are silly, but you seem to be saying that they don't really pick up any useful signal. Do you know that to be true?
> The math portion only tests up to Algebra 2. Many students have taken math beyond Algebra 2, so it doesn't even show colleges what their actual math ability is.
This is more of an argument for the SAT's math section to be made more difficult.
> I think grades are more relevant because they show how students performed across multiple subjects over a longer period of time. Grades are more subjective and possibly gameable. But I don't think ability to answer psychometric questions is a better metric than grades.
I see grades and the SATs as simply being _different_. Neither is strictly better than the other because they aren't really showing the same thing. In that case I see no reason they can't coexist.
You're correct, they are not required. I think that rather than a requirement, students should be able to submit these AP/SAT II scores to add value to their application. Some schools don't offer AP courses, so it can't really be a requirement. SAT II could be made a requirement. My opinion is that if we're going to use standardized tests at all, let's use standardized tests that measure actual knowledge, not "psychometric abilities" (i.e. SAT II and/or AP tests).
>> Do you know that the tests don't really correlate to future success or are you speculating here?
There is a lot of research on SAT/ACT scores and college success. It's a mixed bag. Standardized test scores basically only correlate with grades in the first year of college, and don't correlate at all with graduation rates. GPA, however, does correlate strongly with college success. There are a lot of research studies correlating SAT/ACT scores with different metrics. And it turns out, that a "psychometric abilities" test just isn't a great metric for how people do in college. Why are we investing billions of dollars and hours in a really ineffective metric?
>> This is more of an argument for the SAT's math section to be made more difficult.
I think we should just have math subject tests instead. That way kids who only took up to Algebra II take one test appropriate for their level. And kids who took Calculus take another test appropriate for their level. A history major probably doesn't need calculus. A physics major does.
> I see grades and the SATs as simply being _different_. Neither is strictly better than the other because they aren't really showing the same thing. In that case I see no reason they can't coexist.
Well, research has shown that grades predict college success and ACT/SAT doesn't really. So I would say that one is strictly better than the other. I'm not saying that standardized testing shouldn't exist. I do think students should be able to take standardized subject specific tests like AP and SAT II and submit those scores with their application. Grades and standardized tests can co-exist. Let's just not waste time and money on a standardized test that doesn't provide a meaningful metric.
I don't really take issue with the rest of what you say, but you simply can't claim that grades are strictly better that SATs. They are _different_ measurements. There are people with bad grades that will do well at SATs. Now if by "strictly better", you mean that grades individually are a better predictor than SATs, then that I could believe (though obviously it's a claim that requires actual study), but that isn't really a good definition of strictly better. It could very well be the case that grades provide 50% of the predictive signal of success while SATs provide 35%, but taken together they provide 60%. In that case, neither is really strictly better than the other. However, both together really are strictly than either separately.
That said, if by "research has shown that grades predict college success and ACT/SAT doesn't really" you mean SATs provide 0 predictive power, then they should of course be removed. I've never heard anyone make that claim, but it could be true I guess. I'd have to defer to your knowledge of such research.
But assuming that SATs aren't providing 0 predictive power, I'm not convinced these tests should be replaced with nothing at all. If they replaced them by required SAT subject tests, then that's one thing, but they aren't doing that. They are simply removing a signal. Before they did have a single test to compare across and now they have none.
Anyway I don't really feel like going back and forth forever with this. My original question in this thread about whether the SATs provide information has been answered well enough. Thanks for the responses.
"UChicago Consortium researchers found that the predictive power of GPAs is consistent across high schools—something that did not hold true for test scores. At many high schools, they discovered no connection between students’ ACT scores and eventual college graduation. The authors were also surprised to find that, at some high schools, students with the highest ACT scores were less likely to succeed in college."
Students can already submit other standardized test scores like AP to improve their application. I think those tests a) contain material that is worth the time study, and b) measure something clear (i.e. a student tried to learn chemistry and was able to do so, as measured by this exam. And can use the skills learned for this exam in the future.)
https://news.uchicago.edu/story/test-scores-dont-stack-gpas-...
I would argue that elite schools mostly want students who will become wealthy leaders. The people who achieve this are either already wealthy and connected, or think creatively and can design/implement a unique solution. I don't think the SAT/ACT can give much of a signal on the latter.
The main purpose the SAT has served is making it easier for admissions staff to make and justify decisions.
I think subject-based exams are a better use of students' time, and give a clearer signal to both the student and the school whether they will be successful in their chosen major. (i.e. A declared history major performing well on the AP US History and AP World History exams. A declared physics major performing well on an AP physics, AP statistics, and/or AP calculus exam.)
Also, a student's ability to succeed in subject-specific exams is more dependent upon the HS's offerings. For example, I went to a school that had a very strong science program (e.g., regularly top-3 in national competitions). But if I was trying to prove my ability to be a strong student of social sciences, there were very few courses I could have taken to demonstrate that.
Lastly, most students take a plurality+ of their subject-specific exams in 12th grade, which is too late for them to affect admissions. If more reliance were placed on subject-specific exams, it would advantage kids at competitive high schools, where there are so many APs offered that students take many of them in 10th-11th grade.
>> Also, a student's ability to succeed in subject-specific exams is more dependent upon the HS's offerings.
While some schools don't offer all of these subjects as courses. I don't know of a single school that offers "SAT/ACT" as a course. So while my school may not have all the AP classes, I could select from the classes that are offered and at least take those exams. Students can also take AP exams without taking the course (which is what they already do for the SAT).
>> Lastly, most students take a plurality+ of their subject-specific exams in 12th grade
You are right that many students don't take AP classes at all, or not until 12th grade. In that case, I think SAT II subject tests are preferable to SAT I, because at least students spend time studying a subject that is useful and transferable.
Basically, I think a system of: Pick 1-3 SAT II or AP exams in subjects of the student's choosing to add value to their application would be better than using SAT 1 scores. That way:
-Students show their college readiness by learning transferable knowledge.
-It is standardized so it compares apples to apples.
-Students can learn the material in school, and focus on the subjects that they prefer or are offered in their school.
-If the subject is not offered in their school they can still self study (like they do with SAT 1) and take the exam anyway.
-I think the exams should just be used to assess "college readiness" not what major they can enter. That way students aren't pigeonholed into a specific major.
-I think these exams should be way to add value to a college application rather than a requirement to address the issues of accessibility. If someone went to a high school that offered 0 AP classes, and they don't take these exams I don't think they should be rejected from college. If they take 1 AP exam that they self-studied for, that shows a lot of initiative and could add value to their application.
Also, any argument that favors ignoring SAT/ACT scores would also apply to AP tests. In fact, the barrier to AP tests is much higher. Any kid can sign up for the SAT, get a fee waiver (if applicable), and study with books from the public library. Taking an AP test is much more complicated; I don't know if it's even possible if your school doesn't offer the course.
On the SATs you’re studying how to memorize the ocean. With AP classes you’re taking a “college level” course that allows you to save money by skipping courses. You can’t meaningfully Khan Academy SATs, and yet it is a world class resource for standard K-12 instruction.
At least in my limited experience with the people at the UCs, the AP classes only counted toward credit hours. They did not count towards GE and Major requirements. In that, if you got a 5 on the Physics AP, you still had to take Physics I to get the GE and Major requirements [0].
Also, your score on the AP test was the credit hours you received. So a 3 on the US History AP only gave 3 credit hours, while a 5 on the Chemistry would give 5.
The people I know that went through this had done so a while back. So thing may have changed, of course.
[0] As in, if you needed/wanted to later take Physics II, you had to take Physics I at the UC, the AP test would not waive the requirement.
They must have changed since when I took them. At that time it was a pretty well-defined list of mathematical concepts and language skills. The only memorization I remember was vocabulary, which wasn't ocean-sized by my recollection.
Per the California Master Plan, I thought it was the top 5% for the UCs and the top 15% for the CSUs. And the top percentage was on a school by school basis, not the whole state pooled together.
Meaning that students from very good high schools are looking to have a quite tough time getting in to a UC as compared to students from very bad schools.
Things may have changed as it has been a while since I looked into this.
You can see their reaction to the success of their action here:
https://www.facebook.com/68134933006/posts/10159313850238007...
https://mobile.twitter.com/anilaali/status/13858431187219783... (“Was working on an essay about how ‘Asian American’ is pure politicized astroturf manufactured by the activist class and non-profit industrial complex speaking to no actual ethnic constituency and she just...tweeted it out”).
Asian activists stand in solidarity with other progressives on opposing policing even as a response to growing anti-Asian hate crime. Meanwhile, most Asians in the neighborhoods where they’re getting attacked support the police: https://slate.com/business/2021/04/chinatown-block-watch-kar...
> I’m curious how people in the block watch talk about things like defunding the police. I’ve noticed that some activists are pretty bullish on the idea of taking money away from the police and saying this really isn’t working for our community, while others very much want the protection of the police. What are those conversations like when you’re having them?
> Well, actually, there are no members on the Chinatown Block Watch calling for defunding the police.
> None?
> We don’t get into partisan politics. We actually don’t talk about politics at all. I support community policing here. I would love to see a beat cop come back onto the job. Decades ago, there used to be a beat cop, these officers that walk around the neighborhood on foot, and they get to know all the business owners, they get to know the residents.
That was what also happened with the move to whole language reading away from phonics, which is a tried-and-true system of grapheme-phoneme mapping (and typically followed by word decoding with prefixes, suffixes, and root words), that ruined reading comprehension for several generations; this was also the impetus for the formation of the Challenger private school system. Apparently, the California Department of Education returned to phonics many years later because it was such an utter failure. The problem is these arbitrary experiments are committed-to and continued due to image and political concerns and sunk costs bias, but without any fair and complete validation that they work.
What the heck was the goal of shipping talented kids around to inferior schools where more students had emotional problems, broken homes problems, anger issues, and lower performance? It's not like gifted kids are going to do better in a shittier environment. It seems more like punishment or padding metrics at the expense of students.
Most people aren't really there for a liberal arts education; they're there to try and secure future employment.
Support apprenticeships and on the job training alongside online courses and universally accessible community college.
It won’t happen without a significant cultural shift either. There seem to have been a number of institutions that have attempted to make education affordable and they all fail for various reasons. One being that at the end of the day a college degree is the only sort of credential considered respectable and even those can be worthless depending on a number of factors.
Also, the number's probably more like hundreds of thousands or perhaps millions when you look internationally.
Another issue is that the degree is valuable because it is selective and admits high quality students who will be high quality professionals as well. If a larger population is admitted because they can do the coursework but they are not quite at the same tier of capability, the value of that diploma will reduce.
On your second point: a Berkeley diploma signals that a graduate (1) is above some minimum (but still very high) threshold of competence, and (2) has a level of competence governed by some probability distribution that's above that minimum threshold. I assume that as Berkeley admitted more students, the minimum standard would stay the same but the average of that competence distribution would decrease toward that minimum.
This is subjective and I'd understand if you disagree but that seems like a worthwhile tradeoff to me. I think there are pretty few jobs for which a bottom-of-their-class Berkeley CS grad is underqualified, and obviously those jobs have their own evaluation/screening processes on top of looking at the Berkeley diploma anyway. I suppose an employer looking for the "superstars" in the top 1% (by some immeasurable criteria) of their class at Berkeley would have to identify the top 0.1% instead (if we generously assume that Berkeley's current admissions process never turns down those students), but I'm less concerned about those dozens of students than about the tens of thousands more who could benefit from both a more rigorous curriculum and the ability to demonstrate to employers that they're above that minimum threshold.
that won’t change.
Reason N to not donate to my alma matter.
How can a math test be biased? Please keep it concrete, related to these actual tests, and not reach for hypotheticals
You can consider SAT performance reflects:
A) SAT test prep
B) Other things that only more wealthy/affluent parents can easily provide.
C) Other stuff.
It's possible that A (test prep) has little effect on its own, but that A+B has a massive impact.
The linked article talks only about the impact of test prep. I can't access the cited sources, so don't know whether they've attempted to control for any correlation between those students who received test prep, and those students who received other special advantages that may also have increased their SAT scores.
There are good arguments on both sides.
But, as other commenters have pointed out, the conflict only exists because:
1. We rely on university degrees as a credential to filter people for jobs and other desirable opportunities.
2. To maintain the value (and price) of the credential, the best universities must restrict supply.
3. Outcomes from the USA's public grade/high school systems vary widely, and no one has figured out how to give everyone a good (if not equal) shot.
So perhaps there's no perfect solution possible from tweaking one thing (university undergraduate admissions). The whole system needs work, from K onwards. But such an overhaul would need people to agree on what they're trying to optimize for.
2. The school's credential is valuable because it's a proxy for smart, well educated people. If Harvard stopped looking at grades and sat scores and instead just picked 400 people at random the degree would be worthless.
This would not be the case if either:
A. There are high standards for graduation, OR
B. Attending Harvard for 4 years makes you more capable.
If you're suggesting that 400 applicants picked at random would all graduate, then maybe the standards need to be raised. If the degree signals only that you were good enough to get in, then the 4 year gap between matriculation and graduation seems like waste.
B. I doubt that the Harvard educational material makes you significantly more capable than any equivalent one. Same text books, doubt teachers are that much more capable than a good state school.
If you're a smart and poor kid you're more likely to be undereducated and unprepared for college level academics.
> For the most part, wealthy families who can afford private tutoring for their kids at $150 or more an hour, generally don't send their kids to the Jiechu School for help on the SHSAT. It serves families of modest means, especially recent immigrants, many of them from Fujian province, who constitute the lowest socio-economic layer of the Chinese-American population.
> For them, Jiechu's price – $650 for a seven-week summer course, four hours a day, Monday to Friday – is a doable stretch. That
I had never heard of Jiechu until just now, but you could achieve this 50% cost reduction simply by having class sizes closer to those of China's high schools (~48 per class) than those in the US (~24 per class).
This is before you consider other potential cost savings, e.g.
- contract teachers can be cheaper than unionized employees
- no costly central bureaucracy
Stats are less gameable than alternatives like grades, lacrosse trophies, and interviews/subjective chemistry evaluations.
At the very least they should have kept SAT math and subject SATs for science, engineering and CS admissions. Next are they going to stop considering GPAs and AP scores since I’m sure there is a correlation too.
What is going to be the admission criteria now? Random number generator? ‘n’ from each high school?
Also if you don't have a safe place to study that's awful. But that's going to affect your grades and how much you learned the previous 12 years much more.
It's worth considering that we have been wasting an enormous amount of people's time and energy training them to answer contrived questions that have very little to do with how prepared they are for college and how likely they are to succeed. Just because I had to do it, I see no reason to inflict that BS on today's kids.
And grades are a better of future grades than just an sat score but sat score plus grades are better than just grades.
But I also wonder how well they correlate with job success after school.
100 years ago there were Jewish quotas in Ivy League schools and others. The way that they finally were able to get over the quotas was when universities started accepting the SAT for admission, thus making admissions merit-based.
This is easy history to look up on the web, so the reality is that merit-based exams help to overcome quotas.
I wonder if now the universities will re-instate the quotas for Jews, since Jews are overrepresented in universities compared with the proportion of those that appear in US population.
Any opinions about this?
Is this "SATs and ACTs are racist!"?