Daniel Asarnow
@dabiophysicist.bsky.social
Postdoctoral scholar in Veesler Lab @ UWa.
Former UCSF Biophysics & SFSU & UCSC Biochemistry student & Exploratorium Explainer. He/him.
I take tiny pictures, in 2 & 3 dimensions.
https://github.com/asarnow
Former UCSF Biophysics & SFSU & UCSC Biochemistry student & Exploratorium Explainer. He/him.
I take tiny pictures, in 2 & 3 dimensions.
https://github.com/asarnow
I also think they are bias machines, against those unfamiliar (URM), against careful thinkers & ND (timed), against ELLs (slightly ameliorated by Prop 58), against poor advising (retakes, fee waivers, extra time)...the university was to be "as the public convenience may demand" not the inverse.
December 1, 2025 at 10:59 AM
I also think they are bias machines, against those unfamiliar (URM), against careful thinkers & ND (timed), against ELLs (slightly ameliorated by Prop 58), against poor advising (retakes, fee waivers, extra time)...the university was to be "as the public convenience may demand" not the inverse.
It also proved to be a terrible predictor of performance - I failed calc I and II (for chem/bio) more than once, then finally used my AP score for credit and swapped into the math/eng series (which was easier and better taught).
December 1, 2025 at 10:23 AM
It also proved to be a terrible predictor of performance - I failed calc I and II (for chem/bio) more than once, then finally used my AP score for credit and swapped into the math/eng series (which was easier and better taught).
What I believe is that they are tests of test-taking ability at best, with strategy and luck far more impactful than bonafide review. The point isn't that I think it was easy!
December 1, 2025 at 10:23 AM
What I believe is that they are tests of test-taking ability at best, with strategy and luck far more impactful than bonafide review. The point isn't that I think it was easy!
I promise we can do better by our students, with something more rational and more fair than these tests, if it really is *admissions* that need adjustment and not education
December 1, 2025 at 8:59 AM
I promise we can do better by our students, with something more rational and more fair than these tests, if it really is *admissions* that need adjustment and not education
Btw, in 2004 I myself gained 100 points, from 660 to 760, on a simple SAT II Math rerun. Though tbh I may have buffed up my strategies in an aisle at the local B&N.
December 1, 2025 at 8:59 AM
Btw, in 2004 I myself gained 100 points, from 660 to 760, on a simple SAT II Math rerun. Though tbh I may have buffed up my strategies in an aisle at the local B&N.
You may restrict your claim to a particular band of math-section scores sent specifically to UCSD, etc, etc, but it's inescapable that many students in absolute terms, particularly underrepresented ones, will fall over thresholds because they lack access to broad testing resources.
December 1, 2025 at 8:59 AM
You may restrict your claim to a particular band of math-section scores sent specifically to UCSD, etc, etc, but it's inescapable that many students in absolute terms, particularly underrepresented ones, will fall over thresholds because they lack access to broad testing resources.
Your claim was that even dedicating additional *resources* to students should produce roughly no effect, yet these data show quite consequential changes with nothing more than a second hack. (Likely due to test familiarity).
December 1, 2025 at 8:59 AM
Your claim was that even dedicating additional *resources* to students should produce roughly no effect, yet these data show quite consequential changes with nothing more than a second hack. (Likely due to test familiarity).
The average second take score was 46 points higher than average first take scores, but the average overall *increase* most generally relevant to admissions (superscore) was 90. You cited 50-point bands for math placement levels. Overall swings can distribute many ways ofc.
December 1, 2025 at 8:59 AM
The average second take score was 46 points higher than average first take scores, but the average overall *increase* most generally relevant to admissions (superscore) was 90. You cited 50-point bands for math placement levels. Overall swings can distribute many ways ofc.
I hope you say that regarding only the SAT. FYI the average score increase on a single retest is nearly twice the difference between math placement bands you cited, and boosts 4-year enrollment probability by >10%.
Take Two! SAT Retaking and College Enrollment Gaps
(May 2020) - Only half of SAT-takers retake the exam, with even lower retake rates among low-income students and underrepresented minority (URM) students. We exploit discontinuous jumps in retake prob...
www.aeaweb.org
December 1, 2025 at 6:03 AM
I hope you say that regarding only the SAT. FYI the average score increase on a single retest is nearly twice the difference between math placement bands you cited, and boosts 4-year enrollment probability by >10%.
We haven't touched on potential conflicts of interest regarding sliding financial aid, or the logistics, and consequently equity, of universal SAT reimplementation (as opposed to the previous argument about intrinsic equity), but they are on my mind as well.
December 1, 2025 at 3:32 AM
We haven't touched on potential conflicts of interest regarding sliding financial aid, or the logistics, and consequently equity, of universal SAT reimplementation (as opposed to the previous argument about intrinsic equity), but they are on my mind as well.
To be clear I'm not arguing against any changes! I think ELC could be made mandatory & used to hurt some schools/students instead of helping only, for example (like the report method, but using more of the data available to UC).
December 1, 2025 at 3:32 AM
To be clear I'm not arguing against any changes! I think ELC could be made mandatory & used to hurt some schools/students instead of helping only, for example (like the report method, but using more of the data available to UC).
And yes, I am currently of the opinion that the (purchasable) bad grades/good scores pathway is likely less fair than alternatives and should probably not return. Full disclosure: I used it myself, though I did not pay
December 1, 2025 at 2:51 AM
And yes, I am currently of the opinion that the (purchasable) bad grades/good scores pathway is likely less fair than alternatives and should probably not return. Full disclosure: I used it myself, though I did not pay
I never suggested it should. I am saying that UC should continue to meet its obligations to the state, and that local context data can work both ways. I am noting that AP scores are still used. I am asking why UCSD can pass the buck on its own basic math curriculum, which they admit is not working.
December 1, 2025 at 2:45 AM
I never suggested it should. I am saying that UC should continue to meet its obligations to the state, and that local context data can work both ways. I am noting that AP scores are still used. I am asking why UCSD can pass the buck on its own basic math curriculum, which they admit is not working.
Paying for retesting and paying for test tutoring are highly effective at raising submitted SAT scores. Did you know that correct answers can often be selected without referencing the question text? IME only English proficiency limits teaching these methods (which raises another whole issue...)
December 1, 2025 at 2:14 AM
Paying for retesting and paying for test tutoring are highly effective at raising submitted SAT scores. Did you know that correct answers can often be selected without referencing the question text? IME only English proficiency limits teaching these methods (which raises another whole issue...)
Math *placement*
December 1, 2025 at 2:10 AM
Math *placement*
Ultimately UC is supposed to admit a fixed proportion of students, not students at a fixed level of ability. Beyond that their training program should be able to exert a normalizing influence, which UCSD admits it is failing to do here
December 1, 2025 at 1:58 AM
Ultimately UC is supposed to admit a fixed proportion of students, not students at a fixed level of ability. Beyond that their training program should be able to exert a normalizing influence, which UCSD admits it is failing to do here
The program they're talking about, that admits these students, is based on GPA and class rank at their school, and long predates dropping SAT/ACT. Remember too that they're still using AP exam scores, which are directly comparable to class grades.
December 1, 2025 at 1:58 AM
The program they're talking about, that admits these students, is based on GPA and class rank at their school, and long predates dropping SAT/ACT. Remember too that they're still using AP exam scores, which are directly comparable to class grades.
Why would SAT and ACT tests, already known to be poor predictors of college performance and strongly influenced by economic status, be better than simply using the grade distribution data they already have in a new way, as the UCSD report recommends?
December 1, 2025 at 1:10 AM
Why would SAT and ACT tests, already known to be poor predictors of college performance and strongly influenced by economic status, be better than simply using the grade distribution data they already have in a new way, as the UCSD report recommends?
It would certainly be punitive, as well as unprecedented and likely illegal, to ban students from an otherwise accredited CA high school! That is *very* different from just considering "local context" as is actually already done to admit high ranking students from (generally) lower income schools.
December 1, 2025 at 1:10 AM
It would certainly be punitive, as well as unprecedented and likely illegal, to ban students from an otherwise accredited CA high school! That is *very* different from just considering "local context" as is actually already done to admit high ranking students from (generally) lower income schools.
Would you mind sharing one or two of the games you have in mind? 🤓
December 1, 2025 at 12:47 AM
Would you mind sharing one or two of the games you have in mind? 🤓
Apologies. Indeed no punitive measures against high schools have been suggested. Here you can see per UCSD's student newspaper that the relevant students are purportedly from schools with eg very high rates of free and reduced price lunch eligibility.
Admissions report finds academic preparedness deficiencies in incoming UCSD students
UC San Diego’s record-high enrollment seems to have come at a cost: increasingly underprepared students. On Nov. 6, UCSD’s Senate-Administration Working Group on Admissions released a new report that ...
ucsdguardian.org
November 30, 2025 at 4:21 PM
Apologies. Indeed no punitive measures against high schools have been suggested. Here you can see per UCSD's student newspaper that the relevant students are purportedly from schools with eg very high rates of free and reduced price lunch eligibility.
There's no blacklist, the UCSD report proposes an adjusted math GPA index & sending feedback to schools at issue. There also isn't such a list of select HS - in fact, this debate is about reducing admissions from less successful schools via ELC. Btw they still use the AP test scores in admissions
November 30, 2025 at 3:50 PM
There's no blacklist, the UCSD report proposes an adjusted math GPA index & sending feedback to schools at issue. There also isn't such a list of select HS - in fact, this debate is about reducing admissions from less successful schools via ELC. Btw they still use the AP test scores in admissions
Well, "blacklist" is a provocative term, which UCSD didn't use. Focusing on UC admissions and not HS education also allows some people to suggest UC should reject its mandate to take the top 12.5% of CA HS grads and either take fewer or different students, rehashing bitter disputes of the 00's.
November 30, 2025 at 11:23 AM
Well, "blacklist" is a provocative term, which UCSD didn't use. Focusing on UC admissions and not HS education also allows some people to suggest UC should reject its mandate to take the top 12.5% of CA HS grads and either take fewer or different students, rehashing bitter disputes of the 00's.
The implication is that some admitted students would be denied if SAT/ACT were used (AP still are btw). UC is supposed to admit at least the top 12.5% of expected CA high school grads, which it is currently doing. Who do you think that denies, who should be let in instead?
November 30, 2025 at 11:01 AM
The implication is that some admitted students would be denied if SAT/ACT were used (AP still are btw). UC is supposed to admit at least the top 12.5% of expected CA high school grads, which it is currently doing. Who do you think that denies, who should be let in instead?
UCSD's report only suggests "giving feedback" to supposedly grade-inflating schools. Most CA highschools participate in ELC, which includes fairly extensive data submission on a greater portion of the top students than UC must admit. The background distribution at each school is available to them.
November 30, 2025 at 10:44 AM
UCSD's report only suggests "giving feedback" to supposedly grade-inflating schools. Most CA highschools participate in ELC, which includes fairly extensive data submission on a greater portion of the top students than UC must admit. The background distribution at each school is available to them.