The University of California has slammed the door shut on using any standardized test for admissions decisions, announcing Thursday that faculty could find no alternative exam that would avoid the biased results that led leaders to scrap the SAT last year.
UC Provost Michael Brown declared the end of testing for admissions decisions at a Board of Regents meeting, putting a conclusive end to more than three years of research and debate in the nation’s premier public university system on whether standardized testing does more harm than good when assessing applicants for admission.
The University of California has slammed the door shut on standardized testing for admissions, saying no alternative to the SAT can avoid bias based on race, income.
Not only is that a cop-out, it’s a particularly bad cop-out.
Biased indicators themselves are not a problem if you recognize how they are biased and you correct for that bias. If you are an admissions director who believes that higher parental income students do better on the SAT due to access to test prep, then you can compare SAT scores by grouping similar income ZIP codes together.
The issue is that not a full 100% of the differences are due to just income, as part of the effect is that engaged parents recognize their children’s talent, and then move into high achieving school districts. So a smart admissions director would give only a partial penalty for the higher scores from students in high achieving school districts.
The problem is much bigger when it comes to race, as I don’t think that UC is legally allowed to make that type of mathematical adjustment when it comes to race. So in their desire to preserve holistic admissions at all costs, they discard useful information on students abilities.
I also thought that the SAT would provide useful information on students’ abilities until I saw data from the College Board itself in the book by Paul Tough on inequality in education (I think it was called “the years that matter most” and then renamed “The inequality Machine” after the pandemic for a catchier title I guess). I’ll try to remember the data as accurately as possible because I don’t have the book in front of me. For about 2/3 of students taking the SAT, their scores match up well with their high school grades and rigor. So the SAT does not add a lot more useful information to help colleges know if they’re college ready or not. For the rest of the students with so called discrepant scores, about 1/6 have “inflated” SATs in relation to grades and rigor and the other 1/6 have “deflated”scores. The kids with those deflated scores are usually from a low socioeconomic background, and those with inflated scores come from an affluent or at least not disadvantaged background. There is a correlation with race and gender too but I don’t remember the exact details. So it seems like the SAT seems like a barrier to kids from poor backgrounds (without access to test prep) and a boost to kids from affluent backgrounds (with access to test prep) and not much of a determinant for the majority of kids. So I can see why schools, particularly those that want to actively promote using education as a tool to increase social mobility, might want to move away from these tests.
California teachers will still be prepping kids for the STAR tests.
Jon Boeckenstedt has a good blog post on the UC test blind decision, as well as an analysis for which non-CA colleges enroll the most CA students: It's going to be hard
Some ZIP codes may contain some wealthy families and some poor families. Some people may live in a ZIP code where most others have very different SES than they do.
The same can apply to high schools or school districts.
More likely to be discrepent either way than overall.
Black
>
<
Highlighted
Hispanic
>
<
Highlighted
White
<
>
Other
>
>
Race no response
<
>
PI < $30k
>
<
PI $30-50k
>
<
PI $50-70k
>
<
PI $70-100k
<
<
Less likely to be discrepent either way than overall.
PI > $100k
<
>
PEd no HSD
>
<
PEd HSD
>
<
PEd AA/AS
>
<
PEd BA/BS
<
<
Less likely to be discrepent either way than overall.
PEd grad
<
>
Highlighted
PEd no response
=
>
PI = parental income, PEd = parental education
Gender shows a large difference in that female students are more likely to be HSGPA discrepent, while male students are more likely to be SAT discrepent. In general, non-gender groups containing more advantaged students tend to be more likely to be SAT discrepent and less likely be HSGPA discrepent, while groups containing less advantaged students are more likely to be HSGPA discrepent and less likely to be SAT discrepent. However, Asian students are more likely to be either kind of discrepent, while those with parent income $70-100k or parent education with bachelor’s degree are less likely to be either kind of discrepent.
Didn’t the UCs try something with using zip codes in admissions? I vaguely remember my daughter mentioning this (we have both very high and low incomes in our zip code) but this was pre pandemic and I don’t remember ever hearing or seeing anything about this since.
Never heard of it, but perhaps you may be thinking of Eligibility in Local Context, where a top 9% UC-recalculated HSGPA for a recent previous class at the high school is used as a benchmark for gaining ELC status, which is visible to admission readers and also gives admission to a UC with space available if the student is not admitted to any selected UC (in practice, this UC with space available has generally been UC Merced).
It seems to me that now the only route to let UC admissions committees know that one does well on standardized tests is to make National Merit semifinalist or commendation.
Does anyone doubt that we’ll see more and more students from CA with 4.0 UW GPA? Given the number and competitiveness of CA applicants, would other states have to follow in the same footsteps?
Fact is, most kids are too busy and unmotivated to do any test prep. And many poor districts offer test prep for those who want it. Khan Academy is free online, and a test prep book can be gotten very cheaply.
For those whose SATs don’t correlate with their achievement, the answer becomes obvious if you sit with the student as they do a practice test. The straight A compulsive student from the highly ranked public school may have a low SAT score because they approach the test the same way that they approach their schoolwork - slowly, methodically, thoroughly, compulsively. That student is going to be a straight A student in college, too, despite their low SAT score. The straight A student from the low-standard district where the student who shows up and isn’t disruptive gets an A, has a low SAT score because they really haven’t learned what they need to, to do college level work. That student is going to fail at college, despite every possible support, because they weren’t ready for college. They needed to start at a community college taking remedial level classes, rather than be admitted to a university.
When California banned racial preferences in admissions, the classes quickly sorted themselves into heavily Asian and White for the flagships, and more Black and Hispanic at the 4 yr and 2 yr colleges. By taking out the SAT, one takes out the piece of the puzzle that differentiated between top 5th % of a challenging, demanding high school, vs top 5th % of a low socioeconomic school with very low academic standards, thus allowing the UCs to admit a higher proportion of Black and Hispanic students to the flagship U’s than if they had included the SAT in the criteria.
Yes, it’s unfortunate that willing, hard-working students living in low socioeconomic districts don’t have the same opportunity for high level education as similar students from other backgrounds. But the state’s flagship U’s are not the place for remedial education. This problem has to be fixed at the preschool and K-12 level.
Given that they had rejected the recommendation from their own internal panel that had invested significant effort investigating, and recommended that tests be retained, and came up with the laughable “we’ll just create our own, in a couple of years”, this was the obvious goal.
They should have just announced this years ago and not bothered with the charade.
It’s always easy to kick the can down the road (or up the road in this case). As a large organization the UCs and CSUs are at least trying. It might work and it might not. UT Austin (the state flagship) guarantees admission to the top 5-7% (I don’t remember the exact number these days, it used to be 10%) students from all high schools in Texas not just the ones from affluent areas that you claim to be better prepared for college. They seem to be doing well for their students and affluent California kids are applying for admission based on whatever rankings they are following. I suggest we remain optimistic until further notice.
Colleges do not look directly at the PSAT or care about it, although some indirectly care about it by offering scholarships for PSAT-based status (National Merit, College Board Recognition). UCs and CSUs historically have not offered much or any of these types of scholarships, however.
I guess what I am getting at is that so many people say test scores don’t correlate with a student’s ability, but then they say it’s ok for the PSAT ( and it may involve money towards tuition at some schools). It’s not consistent thinking- and I am not saying this for UCs - this is for all schools.