Stanford, Harvard, Dartmouth, Yale, Penn, Brown, CalTech, JHU, and UT-Austin to Require Standardized Testing for Admissions

Applying to College
1733

Yes, unless application volumes durably go way back down at the most selective colleges (I would bet the opposite, meaning test required may lead to a one-time drop but then it will keep trending up), I suspect they will all be adopting some sort of multi-stage process where the first stage is largely automated with just a quick human quality check.

In that world, it makes sense the more data they can feed into the first-stage review, the better. That doesn’t mean they have to do anything simplistic with that data, indeed in an AI-world they can be more or less arbitrarily complicated depending on their institutional priorities and how many applications they want to consider for full human review. But as long as they identify some cases where test scores would help that process make sure the second stage was getting such applications as appropriate, then that is a good enough reason.

And to the extent they are concerned about interactions with socioeconomic factors, they can just make that part of the treatment in the first stage, and indeed for the second-stage humans as well.

2 Likes
1734

I will repeat - returning to requiring SATs has absolutely nothing to do with increasing the number of accepted students from groups that previously were underrepresented.

SAT scores favor kids from families with high income and with parents who are educated. The convoluted attempts at painting test requirements as a way to increase the number of previously underrepresented groups fail because the data do not support this. No college has demonstrated a decrease in the number of underrepresented group when they go test optional or test blind, and no college/university has demonstrated an increase in the number of admitted students from previously underrepresented groups when they reinstated SAT requirements.

The claim that SAT requirements do not benefit applicants from previously underrepresented group has been supported time and again by the data, so I cannot understand why people keep on repeating the debunked claim that SAT requirements DO benefit underrepresented group.

There are many ways in which SAT requirements benefit the college/university, and many ways in which they benefit students from groups that are already well represented or overrepresented in those colleges and universities. The former is why SAT requirements have been reinstated extensively in private colleges that required them before the Pandemic. The latter is why this is popular among alumni and parents who, historically, are most likely to select these colleges for their kids.

2 Likes
1735

If the overwhelming majority of colleges and universities accept the overwhelming majority of their applicants, is any group really underrepresented?

1736

the soft bigotry of low expectations…

1 Like
1737

That is a bit disingenuous, since we’re talking about the situation in individual colleges, not in colleges as a whole.

Please explain.

1 Like
1738

You’ve mentioned this a few times in different contexts. Is it your contention that as long as applicants from “underrepresented” groups get into any college then they have no reason to complain about admissions at elite colleges? If so, then do you apply this logic to all groups, or just traditionally underrepresented groups? For example, do you apply the same logic regarding “overrepresented” groups at elite schools? After all, all students who are applying to elite schools most probably can find some college that will accept them.

2 Likes
1740

[quote=“MWolf, post:1734, topic:3657570”]
The claim that SAT requirements do not benefit applicants from previously underrepresented group has been supported time and again by the data, so I cannot understand why people keep on repeating the debunked claim that SAT requirements DO benefit underrepresented group.
[/quote]strong text

I guess the below comments by the Yale admissions office are “bunk” in your view?

"First, we found that test scores have continued to predict academic performance in Yale College. Simply put, students with higher scores have been more likely to have higher Yale GPAs, and test scores are the single greatest predictor of a student’s performance in Yale courses in every model we have constructed.

"We also found that students who have been admitted to Yale without test scores have done relatively well in their Yale courses. However, we have further found a statistically significant difference in average GPA between those who applied with and without test scores.

“Yale has now enrolled more than 1,000 undergraduates who did not include scores with their applications. In each of those cases, the admissions committee felt confident that it had evidence of a student’s academic preparation from other components of the application. Our analyses have found that applicants without test scores have been less likely to be admitted; concerningly, this was especially true for applicants from lower-income backgrounds and those attending high schools with fewer college-preparatory courses.”

4 Likes
1741

Reminder that there is only one place on CC where discussions about race in college admission are permitted and that is on the one thread in the political forum.
TIA!

1 Like
1742

It completely depends on how the college decides to use them.

This is an oversimplification, but if a college chooses to treat, say, a 1450 from a disadvantaged applicant as more of an admission plus factor than a 1450 from an advantaged applicant, maybe even more of a plus factor than a 1550 from an advantaged applicant, that could in turn help offset other admissions plus factors correlated with socioeconomic status, such as high grades in very advanced classes at an expensive independent private school.

And that’s more or less how Dartmouth said they intended to use them.

And I’m not aware of data showing that Dartmouth could not choose to use test scores in that way. Of course if it just used test scores without any such adjustment for socioeconomic status and other advantages, we know that would benefit higher socioeconomic status applicants. But if Dartmouth chooses to make such adjustments, then it is entirely up to Dartmouth who this helps and how much.

8 Likes
1743

Did most schools already do that? I think that’s fair. SAT publishes averages for the school, so comparing your performance to the average of your high school isn’t hard.
Also I think more attention needs to be paid to the choice of a major. While a 700+ math score might be desirable for an engineering/STEM kid, I am not sure that matters much for poli scie or history kids.

1744

Most of the more selective private colleges, at least, have long said they evaluate everything in context, including test scores. So yes, the concept is nothing new for such colleges.

Some public colleges have more formulaic approaches where admissions and/or merit can be tied directly to test scores without any sort of adjustment.

Personally, I think both of these systems are serving potentially useful purposes.

Edit: By the way, I want to emphasize I don’t think people have to just take what these colleges say at face value. My point is more basic.

When you combine way more applications than you could possibly accept plus contextual and holistic review, particularly of the black box variety (where if you are rejected you have no idea why), what then happens with the admitted class is basically up to the college. If they want to admit more lower socioeconomic status applicants, they can. If they want to admit fewer, they can. Being test required does not somehow limit their choices, because that is precisely what being contextual and holistic means–they have the freedom to make whatever choices they want to make.

So at the end of the day, whatever sort of class they end up with was basically their choice. That’s not to say I completely disbelieve the argument that maybe they will feel comfortable choosing a few more disadvantaged applicants when they submit 1400s SATs and such. But that won’t be automatic either, it is still their choice.

6 Likes
1745

This means that applicants who did not submit test scores were more likely to be rejected, regardless of all other parts of their application. This also means that low income students who didn’t submit test scores were even more likely to be rejected, regardless of the rest of their application.

That is not “Test Optional”. That is “sort of Test Optional, but more so for applicants from upper income families”.

Yale is saying that these applicants “weren’t admitted”, but they are being disingenuous. Yale decided not to admit these students. Yale decided that the lack of test scores were more of a reason for them to reject low income students than other students.

Yale wasn’t being Test Optional. They were pretending to be test optional while stacking the deck against low income students.

Also - “need blind” is a meaningless, since AOs can see the high school that an applicant attended. An AO pretty much has a good idea what the family income of an applicant is, based on their application. From the performance and AP offerings of a high school, to whether the AO has heard of the high school, to the recommendations, to the ECs, and to essays, the SES and parental education of an applicant is there for the AO to see without much effort. Every AO has a pretty good idea what the SES of an applicant is after they have read the application.

1 Like
1746

But it’s impossible to know if the school prefers kids with scores or if kids with scores also had more/higher AP scores, better ECs… just overall stronger application.

1 Like
1747

"Yale wasn’t being Test Optional. They were pretending to be test optional while stacking the deck against low income students.

“Also - “need blind” is a meaningless, since AOs can see the high school that an applicant attended. An AO pretty much has a good idea what the family income of an applicant is, based on their application. From the performance and AP offerings of a high school, to whether the AO has heard of the high school, to the recommendations, to the ECs, and to essays, the SES and parental education of an applicant is there for the AO to see without much effort. Every AO has a pretty good idea what the SES of an applicant is after they have read the application.”

These are remarkable allegations about the honesty of what Yale is saying. “They were pretending to be test optional while stacking the deck against low income students.” Evidence please? If test optional was a ruse, and Yale could still get away with “stacking the deck”, why bring back test required?

“Need blind is meaningless”? If so, I am surprised there has not been any whistleblower writing about this in the NY Times.

Believe what you want to believe.

1 Like
1748

I wouldn’t say “need blind is meaningless”, but numerous reliable sources have referenced admission readers estimating which applicants they believe are SES disadvantaged at need blind Ivy League colleges. For example, the Harvard lawsuit docs state:

Harvard’s admissions officers do not receive information about family income levels, but are asked to identify disadvantaged students during their review of the file based on information they receive about the high school, neighborhood, or other facts volunteered by the applicant.

The Harvard admission reader procedures mention " if the reader has evidence that the applicant may be from a modest economic background, please check “Yes” under Staff Disadvantaged on the Reader Rating Form" and they mention there is an automated value ranging from 0 to 1 that the computer assigns.

Among Harvard applicants, if the applicant gets the SES disadvantaged flag, it was associated with a significantly increased chance of admission compared to otherwise similar applicants who did not receive the SES disadvantaged flag. The Harvard lawsuit docs include an internal Harvard analysis in which the dean of admissions asked the office of institutional research to review if Harvard’s admission system really is able to increase chance of admission for low income applicants while being “need blind”. The review found the following. “Expected” means their model’s expected chance of admission based on application ratings, scores, hooks, and other factors… without considering the SES disadvantaged flag. At time of lawsuit, Harvard seemed to be applying a general boost for being flagged as likely to be SES disadvantaged, rather than specifically focusing on context of test score or test score adjustment.

Less than $40k Income – Expected 6% admit rate, Actually had 11% admit rate
$40-$80k Income – Expected 8% admit rate, Actually had 11% admit rate
$80k-$120k Income – Expected 9% admit rate, Actually had 9% admit rate
$120k-$160k Income – Expected 10% admit rate, Actually had 10% admit rate
$160k-$200k Income – Expected 11% admit rate, Actually had 10% admit rate
More than $200k Income – Expected 13% admit rate, Actually had 12% admit rate

Of course most colleges have a different admission system than Harvard. Most colleges don’t have a $50 billion endowment like Harvard and need to be more concerned about having few enough low income students and/or insufficient enough FA, such that there is enough revenue coming in from tuition to support operations. I expect most colleges do estimate how much tuition revenue they are going to get after FA expenses when they admit a class, as well as have a good idea what portion of admitted class is going to have markers associated with lower SES such as Pell, first gen, … when they admit a class. I expect this to be true both for test require and test optional, which contributes to why highly selective colleges like Harvard usually had little change in such markers when switching to test optional. If Harvard wants x% of class to be “SES disadvantaged”, Pell, first gen, or similar; then they can admit a class that is similar to target regardless of whether tests are required or not. With a 3% admit rate, there are many qualified applicants in these groups to choose from.

2 Likes
1749

Beyond the SES tells in the app, nearly all highly rejective schools use College Board’s Landscape tool which more directly spells out the applicant (and their high school’s) SES picture based on data across several factors. At the school where I work, the Landscape report is part of each app and as easy to see/analyze as any other application component.

Need blind schools don’t know a given applicants SAI (I’m not aware of any school that is not being truthful about that), but AOs don’t need to know an applicant’s SAI in order to pick out the low SES applicants. I’ll leave it to others to discuss whether that’s need blind or not.

2 Likes
1750

“Need blind” means the AO does not consider the need for financial aid as a negative in deciding whether to admit. The schools that do consider the inability to be full pay as a potential negative are “need considered”. At Harvard, Yale and other well endowed colleges where low SES is a positive, “need preferred” seems to be a better description.

That is what was so baffling about the comment by MWolf that “need blind is meaningless” at Yale.

5 Likes
1751

It also means a full pay student doesn’t have an advantage in the admission decision.

How much a given student’s SES might be part of the admission decision at any of these highly rejective schools is unknown. (Noting that most colleges are need blind, but those that don’t meet full need tend to not be concerned that a given applicant might not be able to afford the school.)

Maybe, but these schools could easily fill 100% of their classes with ‘needy’ students…even more so now that that at some schools, families making up to $200K can qualify for some need based aid. So I wouldn’t call them ‘need preferred.’ I see it more as they are strategically balancing the SES levels of a given class.

1752

“Need blind” in a narrow sense is easy to fulfill, in that the FA need information is just kept out of the admission process. However, many aspects of applicants do have some correlation to FA need, so “need blind” in a broad sense of including correlates to FA need is not really possible for colleges with any selectivity at all (since even basic stats like GPA and SAT/ACT scores have some correlation to FA need, even before going into essays, recommendations, etc. that may indicate SES advantage/disadvantage, etc.).