UC slams the door on standardized admissions tests, nixing any SAT alternative

Applying to College
386

The current UC GPA calculator gives credit for up to 8 high school classes that are designated as UC honors . Are you suggesting that this didn’t exist back in 2010-2012, or do you believe this is not a sufficient measure of rigor? I would agree with you on the latter, but that may penalize schools which have few honors classes .

The Academic Senate voted 51-0 to keep the SAT and ACT for five years during which time an alternative could be explored.

I found that interesting the first time I read it, and I have yet to read that study. But this raises two questions.

The first is, what can an admissions officer do with this information? I mean a UC admissions officer can’t actually say that “we like Palo Alto HS because it has a very high average ACT” and then admit a bunch of students from it, and then reject everyone from a high school in Watts because of its low average ACT. However, UC can say that it likes the strong students from Palo Alto HS and we are admitting them because their ACT scores indicates they can handle the rigor of UC, and while we would love to admit more students from Watts, their ACT scores indicate they may have problems with the rigor of UC.

The second question is what is a high average school ACT a proxy for? Student intelligence? Parental income? School rigor? A community that values education? Tiger parents? Depending upon a person’s view of the answer, they may favor more or fewer students from such a school system.

2 Likes
387

I don’t consider this a good measure of rigor, but more importantly, there is not a separate control for rigor. It is just a single number with an arbitrary and fixed weighting. This is important for a regression analysis since cannot compare 3.8 with honors vs 3.8 without honors and determine how much benefit the honors has in predicting variables vs the 3.8, or how much overlap number of honors has with SAT score. It also cannot place the optimal weighting on number of honors for best prediction. It’s very different from the studies that include a separate control for rigor, which consistently find little benefit to SAT/ACT beyond controls.

They can consider measures of rigor besides just average SAT score of HS attended. For example, they might consider how many AP/IB/college level classes and whether the student was successful in those classes, when making admission decisions. Or they might do a more holistic type evaluation of transcript and estimating rigor in the context of school profile. As noted the Ithaca study did this rather than looking at HS quality variables, and came to the same conclusion.

This is important to review, and many colleges do this via internal studies. For example, what does the regression analysis show if you add controls for the measures of rigor such as the AP/IB/college level classes mentioned above? Does the additional benefit of SAT/ACT remain or was the SAT;s primary benefit more predicting which kids had this type of rigor? What happens if you also control for parents income/occupation, or ECs/awards, or all the other criteria used in test optional admission?

Even if the analysis finds that community/parent/SES related variables are driving a portion of the prediction that are not viable admission criteria, it is still valuable information to know how much of the SAT/ACT predictive benefit occurs because it is a proxy for parent/community/SES variables rather than criteria that the college wants to prioritize in admission. It can also contribute to changes outside of direct admission selection.

For example, the regression analysis might find that being wealthy enough to pay for 4+ years of college was correlated with graduating on college on time, after all controls. And SAT/ACT was correlated with income, so SAT/ACT was also correlated with graduating on time. After discovering this relationship, the college might focus on making the college more affordable to improve graduation rate; rather than focus on admitting kids with higher SAT scores.

Unfortunately the number of public studies that control for internal ratings of applicants in non-stat admission criteria is very limited. Without this information, one can alternatively look at what happened at existing test optional colleges. How do outcomes compare for students who were admitted based on the usual criteria vs kids who were admitted based on all the usual criteria except for no scores? Do you see large differences in GPA, graduation rate, portion switching out of math-heavy majors, portion getting desirable jobs, portion accepted to grad/professional school, …? Answers to these questions have been discussed earlier in the thread. GPA and graduation rate generally show little difference. However, admission to professional schools that emphasize standardized test scores may show larger differences.

388

That is not what I am saying, nor have I ever said that. I said that the policies overwhelmingly favor privilege. That is very different.

Low SES kids can be protected through a whole series of measures including recruiting first gen, URM and low SES kids. The middle class kids are the ones that will get crunched by these policies. They don’t have access to private tutors and friendly grading that wealthy kids have and their schools do not have dozens of activities and clubs to show “leadership”, but they are also not going to qualify for any special consideration since they do not come from disadvantaged backgrounds.

The false choice of tests or greater diversity as justification for anti-test policies hide the real impact of these policies. Comparing a kid in Cupertino or Beverly Hills with one from a truly disadvantaged background to show that tests are biased has always been so ridiculous a comparison as to be disingenuous. We all get that wealthy children have advantages in every single criteria for admission, including testing, compared to those disadvantaged kids, that almost no one, and certainly no one in this thread, is arguing that there should not be some leveling of the playing field. So let’s all agree to stop using this argument as justification of test blind and TO policies.

Middle class kids probably have some access to test prep, but not private tutors or prep schools that the wealthy have. They may or may not have access to competitions and other awards, but certainly not on the same level as the wealthy, and while their high schools will have some AP classes, very few are graduating with 15 APs like many of those who post on this site are claiming their kids have. They are also probably not going to schools that are curving grades above a 3.5. These middle class kids are competing with the kids of the rich for those non-allocated spots on a completely unfair basis, and that SAT or ACT was the one place they could put themselves back in contention.

And the University of California’s response is to take it away.

2 Likes
389

mtmind: Thank you for the clarification.
Edit: The test-blind policy is a tool used by the UC regents to help reach their pre-determined goal of increasing representation of students from the lower SES, but there might be other un-intended and perhaps negative consequences to others, and to the UC system as a whole.

1 Like
390

Just want to make a correction regarding UC GPA calculations. The UC’s use 3 GPA’s in their application review: Unweighted, Capped Weighted (8 semesters of qualifying Honors/AP/IB or DE courses) and Fully Weighted (Unlimited semester of qualifying Honors/AP/IB or DE courses). The Capped weighted is used to determine the UC GPA minimum qualifications for applicants and is the most cited GPA on the UC student profiles. Full rigor is recognized with the Fully Weighted UC GPA.

3 Likes
391

UCs were test blind in the recent 2021 class, so we can review how things actually changed, rather than guess at what theoretically may happen. At each and every campus, the portion of “low income” kids had significant increases from 2019 (pre-COVID) to 2021, when switching from test required to test blind. Specific numbers from UCB and UCLA are below.
The UC system defines “low income” as less than 30th percentile income among CA residents aged 30-65, which is ~$60k. While there may have been ~30% “low income” prior to test optional/bind, that percentage increased in fall 2021, when test blind.

Portion of In State Admitted Freshman who are Low Income
2019 (Test Required) – UCB = 30% low income, ULCA = 30% low income
2021 (Test Blind) – UCB = 39% low income, UCLA = 36% low income

There were similar types of first gen and racial demographic changes, but to a lesser extent at some campuses. Specific numbers are below:

Portion of In State Admitted Freshman who are First Gen
2019 (Test Required) – UCB = 28% first gen, ULCA = 29% first gen
2021 (Test Blind) – UCB = 35% first gen, UCLA = 32% first gen

Portion of In State Admitted Freshman who are URM (Hispanic% + Black%)
2019 (Test Required) – UCB = 26% URM, ULCA = 29% URM
2021 (Test Blind) – UCB = 34% URM, UCLA = 33% URM

As noted earlier, public school kids had an increase in 2020 (test optional) over previous test required years and private school kids had a decrease. However, I haven’t seen 2021 stats for this statistic.

The groups above appear to be winners as whole in the UC test blind policy (<30th percentile income, first gen, URM, attends public HS). These are the same groups whose SAT score tends to be lower than typical for applicants with similar rest of application – groups for which the SAT is more likely to be a weak point in the application.

However, it’s not as explicit to whether the losers were more middle income families or high income families. I expect that the losers were more high income than middle income, as high income kids are the ones are mostly likely to have a SAT score that is higher than typical for applicants with similar rest of the application. They are the ones who are most likely to get a boost from having their higher than predicted from rest of application SAT score being considered, and most likely to see a loss when their higher than expected SAT score is removed. However, I haven’t seen relevant stats for the fall 2021 test blind class yet. I expect such stats will be available in 2022.

392

Very late to this. Has the discussion come up yet that because they will continue to be test blind it just drives the number if applications up which is just millions in application fees of rejected students? I remember at a presentation a couple of years ago it was pointed out that UCLA made over 5 million in rejected applications that year.

1 Like
393

If correlation = causation, can we hold test blind and TO policies responsible for some very bad trends in college applications and admissions?

For example, applications at lower tier and open admissions colleges are actually dropping significantly, and many colleges are going out of business. It is irrefutable that one of the reasons for TO and test blind policies is to drive more applications and make top schools seem more selective. This also has the effect of creating a “T20 or bust” mentality among high school students, with many of them questioning the value of attending a college if they can not get into a top, competitive school.

So if some are going to give test blind credit for increasing URM admittances by 4 percentage points at UCLA, should we also hold it responsible for dozens of local schools closing and reduced enrollment at community colleges nationwide? What about tens of thousands of people not bothering to attend college? These schools were critical to the overall higher education system, and the only way to attend college for many working class people, including many URMs, trying to improve their situation.

Interest Surges in Top Colleges, While Struggling Ones Scrape for Applicants - The New York Times (nytimes.com)

College enrollment is down — and dropping - POLITICO

There is such a focus on what is best for the “bragging rights” schools that I think we don’t appreciate the impact that these policies on the broader education system.

394

That’s an interesting point of view. I was thinking about that while following this thread. We talk about the UCs as if they are a monolith but what most people are agonizing over is whether their subgroup will be favored or not in the newest iteration of holistic admissions to UCLA and Berkeley. No one is clamoring to get into Merced.

As for the rest of the colleges suffering, I think that has more to do with cost of attendance than anything else. Parents, when they have the money, are paying quite a lot in comparison to what they themselves paid for education so if they’re going to splurge it better be for a splurge worthy name. The parents that don’t have the money is a different story. One of my husband’s best PhD students went to two community colleges before transferring to a UC (I think it was Irvine) for undergrad. Single mother household and he worked since he was 16 to make ends meet. There was no way he was going to get into a UC as a freshman but he wanted the education and he was able to get it. Was it easy for him? Was it fair? Would he have been able to do things differently now? Does it matter? I don’t know.

2 Likes
395

Some conclusions are straightforward, some are more complex For example, as we saw in the NACAC, Ithaca, Bates, and other linked studies in this thread; test optional kids consistently average lower income, more first gen, and more URM than test submitter kids. These same groups average lower scores compared to rest of application, so this conclusion is straightforward and expected.

  • Test optional applicants are more likely to be lower income, first gen, and URM

We see that test optional colleges have increased enrollment of these same groups that are overrepresented as test optional applicants/admits, so it doesn’t take a lot of steps to draw a logical conclusion about the relationship. If you want to review further, the Ithaca study included a regression analysis with more controls that came to the same conclusion – a test optional admission policy favors admission of groups who are more likely to apply and be admitted test optional, which include groups who average lower relative test scores compared to rest of application, such as lower income.

There are possible exceptions to these generalizations. For example, a need aware college might admit however many full pay kids they need to meet their tuition revenue goals. However, I don’t have any reason to expect UCs to be such an exception, particularly considering that the expected relationship occurred on their first year of being test blind.

I hit a paywall with the NYT article, so I read the Politico one instead. The linked report states the following. Blaming test optional for a 2.2% decrease in bachelor’s degree enrollment or 10.9% decrease in associates degree enrollment is not as logical or straightforward a conclusion. It is not surprising that the article does not do this and instead suggests other explanations.

  • Associate degree enrollment saw a 10.9% drop and bachelor’s degree enrollment declined 2.2%. In contrast, master’s and doctoral degree enrollments are up 5.2% and 3.6%, respectively

.

396

When UCs remove the testing requirement, they don’t replace it with another criterion, do they? If the remaining criteria are weighted proportionally the same as before, the only change is the elimination, not any replacement, of the testing requirement.

397

While they are not replacing SAT with a new and original criteria, the relative weight of the non-SAT criteria that is considered increases. If you think of it as a formula (I realize admission is actually holistic and not a formula) where the sum of the weights has to add up to 100%, the relative weights change when any criteria is removed such that the sum is still 100%. For example, if a hypothetical college has a formula of 50% GPA + 50% SAT = selection rank; then with SAT removed the formula becomes 100% GPA = selection rank – the relative weight given to GPA increases with SAT removed from equation.

For the purposes of the UC’s holistic admission, the admission officers consider 13 types of criteria in assigning their overall score , rather than 14 types of criteria. At least some of the remaining 13 criteria have increased relative weight higher than they previously did with scores required. Whether test blind helps/hurts an applicant depends on whether the relative stronger/weaker part of their application is the combination of 13 remaining criteria, or the 14th removed criteria (test scores).

398

Of course, the remaining criteria are weighted more heavily. But if the increases in the remaining criteria are proportional, it doesn’t matter. It’s exactly the same as a simple elimination.

399

Okay. Maybe you have an issue with the specific wording used with “replace” rather than increase weighting of existing criteria that may remain in proportion to make up for the removed criteria. In any case, the overall conclusion remains the same about which types of applicants are helped/harmed by a test optional/blind admission system, as stated in previous posts.

400

The point of my original post you responded to is that the elimination (or relaxation) of any admission criterion/impediment, not just testing, will make the student body more representative of the population. If that’s the ultimate goal, why do we not eliminate any of the other criteria?

401

No, removal of any admission criteria will make the student body more representative of the groups that do best in the remaining criteria that is used in evaluation; which may be make the admitted class increasingly different from the overall population.

For example, suppose a hypothetical college removed all admission criteria except legacy. The admitted students would all be legacies. That wouldn’t make them any more representative of the existing population than the usual admission criteria. Suppose they instead removed all admission criteria except national level academic awards. That also wouldn’t help make the admitted class look the overall population. If they removed all criteria except SAT score, such that nearly every admitted student had a ~1580+ SAT score, that also wouldn’t help make the admitted class more representative of the general population than their previous admission system.

2 Likes
402

No, that’s incorrect, if the criterion that’s removed is a greater impediment to a segment of the population. For low-SES students, any one of the admission criteria (other than hooks and tips, which help boost admissions of certain groups but aren’t really criteria) is an impediment.

403

I gave an example of eliminating all criteria except for SAT score, such that all students have 1580+ SAT. Do you think that this only SAT admission system would improve enrollment of low SES students because all the removed non-score criteria are impediments?

1 Like
404

It depends on which one is the greater impediment to low SES students. Do we know tests are greater impediments than all the rest of the criteria? If we eliminate all criteria other than ECs, would low SES students be better represented? We know for certain that if we eliminate all of the criteria, they would.

405

If the weight of the SAT is equally distributed to the remaining criteria, then removing SAT might have a small impact. If it has low/no impact, then why remove it?

Clearly, removing standardized tests serves some purpose to make the student body what the school administration wants it to be, right?

I think Asian parents see it as a way to reduce the number of Asians in UC. @CTDad-classof2022 seems to think it will make it worse for SES students.

It seems like a lot of guesswork without insights into the admission process at UC. We will see how it pans out.

1 Like