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

Applying to College
61

The problem that the UC system has with standardized tests is exactly that - they are standardized. That’s where they come in handy, when you are comparing the top students at one school against the top students at another school. It becomes pretty obvious when you see that the top students at an inner city non-selective high school are getting sub 1000 SAT scores, that they are not as prepared for college as the top students at good high schools who have >1500 SAT scores. They’re also useful for standardizing among applicants from the most prestigious schools, and just the best students at good public high schools. You should have seen the pitchforks and torches come out when I wrote on the Prep Schools website that it was obvious that there was grade inflation at some of the most prestigious prep schools, and that their students, on average, seemed to score no better on standardized tests than did the honors/AP cohort (top quarter or so) of a good suburban public high school. “Those tests cannot possibly measure how wonderful Buffy and Chip are!” Well, they can, and they do. That’s the problem.

When you look at the general scores of racial groups, certain groups tend to score higher than other groups. It’s not that the tests are biased, it’s that certain groups tend to be better academically prepared than other groups, and hence do better on the tests! Eliminating those standardized tests is simply a way of making it more difficult for the higher achievers to stand out, thus making it easier for the UCs to admit a more racially balanced class than the ones they admitted when they considered the SAT. Imagine if coaches choosing their basketball recruits could not consider height, or the athletic record of the applicants. They just had to go on the recommendation of the high school’s coach as to how good a player the kid was. It certainly would level the field for all high school basketball players applying for athletic scholarships. Going test-blind accomplishes the same thing. It levels the academic playing field, making the best student at the worst high school appear equal to the best student at the best high school.

The end result is either dropping the academic standards at the top institutions to allow those students who were not ready for the work to pass, or acknowledging that a significant proportion of those students admitted from the top of the class at a low-standards high school will wind up failing out, despite every possible academic support. It’s a disservice to those students, who probably would have done just fine at a less-competitive 4 yr UC public college. But it does achieve the goal of racially balancing the schools, without considering race, per se - just proxies for race, such as geography and socioeconomic class.

3 Likes
62

Not sure what you mean by “activists,” but if it refers to those concerned with education, then I think you are making assumptions that don’t necessarily hold true. Educators at all levels are concerned with the growing disparity in participation and performance between males and females in education, and many are trying to figure out how best to address it. This includes California academics and educators all levels, but unfortunately, due the 1996 anti-affirmative action proposition, the hands of public educators are somewhat tied. (It is fascinating to see how views change on this issues when such things adversely impact males.)

Again, I disagree with your assumptions here. UC administrators aren’t blind. They see what is happening in education between the genders and they do not like it, but again, their hands are somewhat tied by the anti-affirmative action proposition. My guess is they would be thrilled to bring in educators who could actually make headway in this regard without violating the law.

Cancelled permanently? That’s a bit much isn’t it? What does that even mean? Is there any factual basis for you assertion?

Again, any UC administrator would candidly acknowledge they are troubled by the disparities, but the anti-affirmation laws have tied their hands.

63

Pretty sure that this thread has hit on every SAT/ACT trope out there. Here are the top 9 per Jon Boeckenstedt, current Oregon State Provost of Enrollment, formerly of DePaul (both schools went test optional under his leadership and a noted leader in the industry):

  1. The tests are standardized and thus fair because everyone takes the same exam.
  2. The “Diamond in the Rough” (DITR) theory.
  3. Grade inflation makes grades meaningless, and thus we need tests.
  4. Don’t blame the tests for telling the cold hard truth.
  5. Removing the tests doesn’t accomplish the diversity objectives that are often an explicit or implicit goal of going test optional.
  6. More information is better.
  7. Everything else we use is bad, so we need tests.
  8. Tests measure native ability.
  9. If low-income students don’t take a test because they don’t need it for college admission, colleges won’t be able to find them.

I don’t understand why any non-enrollment management professional on these boards thinks they know better than the experts who read thousands of applications per year. If you listen, really listen, to what many of the enrollment management industry leaders are saying (people like Jon Boeckenstedt, Rick Clark, Whitney Soule, etc.) you hear them say again and again that they don’t need test scores to identify students who will succeed in college. Why can’t/won’t posters on this thread believe that?

Here’s Jon’s longer post on the tropes: Some final thoughts on the SAT and ACT – Jon Boeckenstedt's Admissions Weblog

12 Likes
64

Unfortunately from this point of view, colleges over the years/decades generally have been moving away from the SAT subject (formerly Achievement) tests that were actually intended to measure achievement in the associated high school courses, so they fell into disuse by most colleges and were discontinued this year, probably because the College Board felt that the market was too small.

Of course, AP tests remain, but they are only useful for admission purposes when students take AP level courses before 12th grade*, which is probably not a large consistent sample of a common set of AP tests even among those students advanced enough to be taking some before 12th grade.

*Unless colleges go to a UK-like system where admission is contingent on matching predicted AP scores on 12th grade AP tests. But even then, moderately and less selective colleges may not have a lot of AP courses / tests taken among their students.

1 Like
65

Students’ test scores (along with all the other academic indicators) are more correlated with educational attainment of both their parents than anything else, including family wealth. Perhaps keeping highly educated students from marrying each other is the ultimate strategy to help achieve equality. :stuck_out_tongue_winking_eye: :laughing:

66

There’s opposition to all kinds of testing, not just standardized tests. If you’re opposed to comparing students from one school to another, you’re probably against comparing students within the same school or in the same class. Imposition of any kind of test is never going to be popular, not on CC, not in the general public.

67

"…enrollment management industry leaders are saying… again and again that they don’t need test scores to identify students who will succeed in college. "

IMO, that is a mighty low bar.

1 Like
68

How so? Many colleges want a well rounded class, not a class full of students with straight As, 98%ile test scores, high rigor, etc…and we hear the admissions personnel at many schools say this.

We can also see this in the admissions results at the highly rejective schools, where many students with very high academic stats do not receive admission and students with lower academic stats do. All admitted students fulfill some institutional priority of the college, and we know institutional priorities go beyond academic stats. Institutional priorities can change from year to year too. Building a class to fit whatever factors a given school desires is their prerogative. Seems simple.

4 Likes
69

Tests can serve purposes other than ranking students against each other. Education isn’t necessarily a competition between students. Not sure why that is so hard to understand.

Perhaps it depends on whether one views college as educational experience or an award for test performance.

4 Likes
70

That’s not a good example of how test blind works. Many studies have found that information gained from test scores is largely duplicated in other components of the application. For example, in several previous posts, I’ve referenced the Ithaca test optional study which found that:

HS GPA + SAT + Rigor + AP hours + … – Explains 44% of variance in cumulative GPA
All of Above with SAT Removed – Explains 43% of variance in cumulative GPA

SAT score did add significantly to prediction of cumulative graduating college GPA at Ithaca beyond HS GPA in isolation, but it did not add significantly beyond the combination of HS GPA + rigor + AP hours + and other factors. The more additional factors you consider, the less scores add to the combination of remaining factors. This relates to how highly selective private colleges like Caltech are able to create a quality class while continuing to remain test blind. Caltech considers many factors beyond just HS GPA in isolation, which allows Caltech to better distinguish being different applicants with similar HS GPAs. I’d make a similar statement about the many other highly selective colleges which were test optional during COVID. They consider rigor, whether classes with non-A grades are relevant to planned field of study, LORs, essays, ECs/awards, interview, … many factors beyond just HS GPA in isolation.

UC differs from Caltech and most highly private selective colleges in that they consider fewer of these additional factors. Furthermore a significant portion of kids attending UCs are primarily admitted through class rank alone. This ELC admission by rank is a UC policy that existed prior to going test blind. This lack of additional factors may lead to admitting some groups of students who are expected to have a lower cumulative graduating GPA average than other potential admits who might excel more in additional criteria that was not considered, which includes but is not limited to scores.

This partially relates to UC’s primary goal is not to admit the class who has the highest possible cumulative GPA prior to effects of a curve. UC wants kids who will graduate, but it would not be a disaster if the average graduating GPA was 3.4 instead of a potential 3.5 prior to effects of a curve. UC has many additional goals in choosing their class. For example, one of the UCLAs admission criteria listed on their website is quoted below. UCLA might favor admitting a kid with a slightly lower predicted college GPA who might make the college a better place while attending.

Likely contributions to the intellectual and cultural vitality of the campus.** We look for a broad range of intellectual interests and achievements. We value evidence of ability and desire to contribute to the campus, and to its diversity: cultural, socioeconomic, and intellectual. We consider the likelihood that a student will interact socially and intellectually with faculty and fellow students, inside and outside the classroom.

Along the same lines, one of the 13 official criteria UC considers in admission decisions is quoted below. It mentions favoring kids with lower income, first gen, difficult personal/family situation, refugee, etc. UC might choose a kid from one of these groups with lower stats over a typical, high SES kid with better stats.

Academic accomplishments in light of your life experiences and special circumstances, including but not limited to: disabilities, low family income, first generation to attend college, need to work, disadvantaged social or educational environment, difficult personal and family situations or circumstances, refugee status or veteran status.

UC and many public institutions face pressure to have a good representation of the state population, rather than primarily admit high SES, ORM kids from a small number of HSs in wealthy areas. Programs like the previously referenced near auto admit for top 9% rank ELC support admitting kids from a wider variety of HSs, with a wider variety of backgrounds.

1 Like
71

It seems inevitable that this policy will be statistically homogenizing across the UCs. For example, students with (or what would have been with) an SAT score of, say, 1220 will become more common at both UCB and UCM.

1 Like
72

Essays have already been proven to be even more strongly correlated with parental SES than standardised tests. Anybody have data on how strongly rigour is correlated? I’d be shocked, shocked! if if it turned out to be any less correlated with parental SES, and average school SES, too. Same for LOR.

With ECs, there is a possibility of correction in holistic admission by considering family responsibilities and paying jobs as equivalent to the kind of internships and other opportunities mediated by parental SES (just read Cal Newport’s book in how to be a high school superstar - those kids just happen to run into professors at neighbourhood parties or business mentors on golf courses). The question is whether it’s happening.

Standardised tests, no matter their actual shortcomings, were always a convenient straw man.

4 Likes
73

I know you don’t like competitions. What would you call admissions at highly selective/rejective colleges? Are they not competitive? Competitions aren’t always bad, and in fact, they’re often necessary. Competitions are what make economies work, for example. What’s needed is a set of rules and constraints that promote fair competitions.

Tests, first of all, serve the purpose of discovering difficiencies in what one learned in classes or over the course of one’s education. A final exam in a class, for example, identifies those deficiencies but is often the biggest contributor to the student’s grade in the class, which informs not only the student but also the instructors of higher level courses whether the student is ready for them. Standardized tests can also serve a similar role in informing the colleges of the student’s readiness.

4 Likes
74

Just so I understand this discussion better, does the anti-test contingent believe an applicant with a 3.75 gpa and 1100 SATs is more deserving of admission than an applicant with a 3.70 gpa and 1500 SATs, assuming all the other criteria are similar?

75

Caltech isn’t a good example of whether test blind policies will work elsewhere (Caltech is still experimenting, BTW, but I expect it will continue the policy beyond 2023). Caltech has a lot more information (than UCs, for example) about an applicant that makes SAT/ACT test scores redundant. Applicants to Caltech are much more self-selective. They likely have taken many APs (or similar) courses and scored 5s by the time they apply (even though Caltech gives no credit for any of these courses), not to mention academic ECs that can demonstrate their abilities. In other words, that fact that Caltech can go test blind doesn’t mean other colleges should or could.

This is similar to how admissions to PhD programs in AI (the most selective PhD programs by far) work. The most desirable/competitive programs in AI are test blind (a few test optional), but the next tier almost always require test scores. The reason is that those most competitve programs practically require applicants to have extensive research experiences (often with publications) by the time they apply, enthusiastic recommendations, and impeccable transcripts with relevant advanced courses in the subject areas.

2 Likes
76

I love competitions. But not when I’m reading.

77

The only thing in that formula that is standardized would be AP SCORES - but that’s not in the formula, because AP tests are standardized tests (not to mention that most of them are taken in senior year, so the scores aren’t available for college applications for most students).

In my diverse district, marginal students are encouraged to try at least one AP class in senior year, partly because high school rankings include a score for the percentage of students who ATTEMPT at least one AP class. That AP class currently is Psychology, since it has a cool teacher of color, and is considered to be an “easy” AP. Surprise! My kid said that many of the students in that class simply refused to do the readings and the work. It’s the only AP class at the school with a high proportion of "1"s on the test. AP class participation is not an absolute measure of achievement or ability. AP SCORES are, but not just sitting in the classroom.

1 Like
78

I agree many schools have specific goals to admit more Pell eligible students (but these students can’t afford many 4 year schools that don’t meet full need) and more URMs.

Not all the UCs have been successful in enrolling classes that are a good representation of the state in terms of race. For example, UCB’s freshman class (class of 2025) is around 41% Asian. This doesn’t include international students, of which some proportion of that additional 15.2% of the class are Asian….I wouldn’t be surprised if greater than 50% of the class were Asian after accounting for internationals.

The proportion of Asian K-12 students in CA is 12%, so that 40+% Asian at UCB is a significant overrepresentation. This comes mostly at the expense of Hispanic students who are 55% of the K-12 population, but only around 23% of UCB Class of 2025.

UCB data by race for class of 2025:

https://opa.berkeley.edu/uc-berkeley-fall-enrollment-data-new-undergraduates

CA K-12 population by race cited in many articles about the ethnic studies curriculum mandate. Not sure if the numbers presented are all K-12 students, or just public school. https://www.educationnext.org/ethnic-studies-california-unsteady-jump-from-college-campuses-to-k-12-classrooms/

1 Like
79

Yes, especially if that 3.75 was earned at a high school with very low standards, where a student earns an A for showing up and not being disruptive. That student has clearly earned the spot, as opposed to the student from the highly competitive school district where an A requires grade-appropriate reading, writing, and mathematical skills, with all reading and homework completed to exacting standards, plus high scores on challenging in-class assignments, often in classes that are at a higher level than first year classes at most flagship state U’s.

The B+ students at my local high school (which is 40% Black and Hispanic, socioeconomically diverse, and has many immigrant students, coming from households with over 60 different home languages) said that they weren’t challenged until junior year at their flagship state U’s. They said that the honors and AP classes at our public high school were more challenging than the first two years of college, at two excellent, highly-ranked flagship state U’s in the Northeast. I can attest that the nearby highly ranked suburban districts have the same level of rigor and competition in their honors and AP classes, as does our district. And that the level at the nearby inner-city schools, or burnt-out factory town schools with a huge proportion of cheap tenement rental housing, is far, far lower. Where the top 10% from each district will differ, is in their SAT and AP scores. But with the scores eliminated or coming too late for applications, the playing field is leveled. I am sure that the instructors at all highly competitive colleges that went test-optional, or test-blind, have some interesting observations that they could make - if only they dared to speak up.

1 Like
80

I don’t disagree there are likely some interesting observations, but not all would be negative.

Many schools that were test optional pre-pandemic track outcomes of test submitters vs those who applied test optional. Among selective colleges I know Bowdoin and Bates do but they haven’t published data AFAIK (but both happily remain TO, as Bowdoin has been for 50+ years).

Some less selective colleges track outcomes between the two groups as well, including Ithaca and DePaul, both of which have published data showing basically no difference in outcomes such as college GPA and graduation rates among the two groups.

4 Likes