The Misguided War on the SAT

The ‘very selective’ schools overall are successful at educating and graduating students of their choice, which seems ok to me.

In theory, I don’t see a problem with private institutions deciding how they conduct admissions as long as they don’t receive US funding or tax-exemptions for it.

But I’m not sure the American tax payer should be funding private institutions that don’t use some sort of objective comparable metric for admissions at least in some way in their process.

Otherwise, without the ACT/SAT, how can people be sure that universities are at least being objective in the process?

2 Likes

This thread seems to go in circles and is way too focused on what the Ivies think and do. Those students who can’t afford Ivies, having family incomes too high to get decent aid, just cross Ivies off their list. Students are finding very good schools with less historical baggage that are affordable for which they have a choice to submit or not. A 4 GPA is still impressive and schools are perfectly capable at looking at courses taken and ECs. The test optional debate is an obsession at the Ivy level because there is some misguided notion that scoring a 1550 makes you more worthy of an Ivy education than a 1350. How many questions difference on the SAT is that spread really? THink about it.

5 Likes

For many of the highly rejectives which this thread continues to focus on, most admitted and enrolled students submitted a test score.

There are additional objective measures beyond test scores (if one actually believes the SAT is ‘objective’). Classes and rigor are objective. Grades in context of the cohort are objective. Generally LoRs are objective.

As a tax payer in the US I am fine that these schools receive federal money to support pell grant students to take one example. I am also ok with them receiving Federal money for their many research endeavors.

For many of the highly rejectives which this thread continues to focus on, most admitted and enrolled students submitted a test score.

Most? Could you give an exact or rough percentage of people who submit test scores?

And what about the percentage of people who don’t submit test scores?

Grades in context of the cohort are objective.

That’s not objective. Cohorts are not of the same strength - someone who is ranked here is not necessarily of the same ability as someone at another school.

Generally LoRs are objective.

Highly dependent on individual teachers and the quality of school you go to - this is what the Chetty study showed to some extent from my understanding. The quality of LoRs is not going to be objective at all.

Classes and rigor are objective.

Highly subjective and dependent on school rigor unless you’re talking about AP?

And highly rejective colleges don’t mandate that AP tests be taken either.

As a tax payer in the US I am fine that these schools receive federal money to support pell grant students to take one example. I am also ok with them receiving Federal money for their many research endeavors.

Which is a view I’m somewhat skeptical is shared by the population as a whole considering how low trust in higher education is.

How can someone who is worried about religious discrimination be confident that schools aren’t dropping test scores to keep certain religious groups out of college?

It’s not like the average American who is genuinely concerned about this as an issue can be reassured by the dropping of test scores.

Test scores provide a defense against that argument because it sounds like there are increasingly larger numbers of calls for institutions to have their funding cut for this exact reason.

That info is in each school’s Common Data Sets.
For Yale class of 2026, 59% of matriculants submitted SAT, 29% ACT. Probably a bit of overlap across those groups.

Do you think these schools have the wrong set of students or mix or types of students in their classes? If not, what is the issue?

That info is in each school’s Common Data Sets.

It doesn’t tell you the percentage that submitted a test of any kind as a percentage of the class because the figures in the CDS could have significant overlap, which is why I was curious to see whether you had more info.

Because if 59% of matriculants submitted the SAT and 29% submitted the ACT with significant overlap, it means it could be upwards of 30% of a class not submitting a test score which is not small either.

I don’t know what to think which is the point I was making.

How can someone who is deeply concerned about colleges keeping groups out of college be reassured by test optionality?

How can one be certain that schools didn’t implement test optionality because of the outperformance of a religious group?

After all, there’s a current lawsuit that alleges Harvard has been keeping a certain religious group out of college and a current congressional investigation into the issue at the moment.

If schools are dropping test scores to practice any kind of discrimination, this would be very concerning to me.

The point I’m making is test scores provide reassurance to people who will be rightfully concerned about this - there wouldn’t be a congressional investigation if this wasn’t worrying people.

2 Likes

https://www.realclearscience.com/blog/2024/01/23/why_college_students_average_iq_has_fallen_17_points_since_1939_1006608.html

Just saw this article about the decline in college IQ average as colleges have expanded enrollment. IQ is correlated with SAT scores (or at least it was once). Not surprising in a way. One thing that stuck out was that the IQs which were average had only a 50% college completion rate. So IQ=SAT =college success at least on some level. It seems that SAT/IQ is useful information

2 Likes

“The decline in students’ IQ is a necessary consequence of increasing educational attainment over the last 80 years,” the researchers commented. “Today, graduating from university is more common than completing high school in the 1940s.”

IQ has fallen as college access has expanded. It is no longer the realm of the elite, but rather, kids of all backgrounds and abilities are not being encouraged to pursue college. Whether or not that is a good thing is debatable and obviously outside the scope of this thread. However, it has little to do with test optional policy and this trend preceded test optional conditions by a number of years.

The point is that it may be useful to know a measurement associated with college completion. IQ seems associated with that.

The article’s saying that IQ of average college student has declined as enrollments have grown as a percentage of the population seems obvious. What else would you expect? What struck me however was the clear correlation between IQ and college completion

I had to step away so I could catch a couple of HS games, followed by some nfl games (go niners!) with some gutter cleaning tossed in here and there. And the server was down. Overall a good weekend. :slight_smile:

But I also had some time to read the “Yale Admissions Director Favors Submitting Scores” thread, which I did for the first time even though I’ve listened to the Admissions Beat podcast episode much earlier. I realized that that thread pretty much covers most of the topics in this thread other than the NYT article itself, and that the article was really just claiming that the scope of schools where testing provides sufficient utility should be expanded. It appears to me that the opinions and realities of this thread can be usefully organized using this “scope of schools” idea.

At an institutional level of granularity, the admission offices of Brown, Dartmouth, Yale, and MIT (plus a few other very selective schools) have made their conclusions known that they find sufficient utility in the tests. Most made it clear that the decision applies to their particular institution. Noteworthy here is Caltech, whose test blind choice was well explained in this thread by ucbalumnus (and perhaps others) by Caltech’s need for criteria with a higher ceiling than the SAT or ACT provides. The podcast linked by beebee3 was very useful in highlighting how they achieve this, part of which was leveraging input from and participation of professors in the admissions process. They really aren’t on the other side of the spectrum from MIT, but rather further along the same side beyond where the SAT/ACT can help.

At the “Ivy-plus” level of granularity, the various Opportunity Insights researchers have made their findings clear. Deming and Friedman (and perhaps Chetty as well) have made it abundantly clear that the scope of their research is the Ivy plus schools (Ivy schools alongside Stanford, MIT, Duke, and UChicago), and that it did indeed indicate that testing is a better predictor for those schools. An IHE article quotes Friedman: “I think since the pandemic we’ve learned a lot more about how the test-optional policies are operating in practice, with data to support, instead of just anecdotally,” said John Friedman, co-founder of Opportunity Insights and the lead researcher on the study, which links test scores to academic success at selective institutions. “And what we’re learning is, without the test scores, there’s a tremendous amount of uncertainty about whether that student is really at the level that [highly selective colleges] require.” (IHE article 1/17/2024)

However, Friedman does also point out that “in many cases” GPA is a better predictor in “more open-access institutions.” Deming echoes these views.

Using an even wider scope of schools, we see the landscape change. Excluding Ivy-plus (and some other very selectives), the situation shifts more to what pilate noted in his posts, that his data points to SAT/ACT as a less predictive force (unless he restricts the data). The third person on the Admissions Beat podcast episode, the VP for undergrad admissions at Clark, Emily Roper-Doten, articulately emphasized this point as well (see below). And of course the important work done by Bowdoin, Bates, Depaul/Ithaca, etc. (why the changes in the utility of testing occurs is interesting to me, but alas not the topic at hand).

Now, there is a little interesting dependency here on majors, more specifically STEM majors. Yikkblue has noted personal and anecdotal reports of students who are less prepared in math… and my understanding is that this is not a rare phenomena (but perhaps not consistently prevalent)… this is corroborated (but not proven) by the relatively higher call for scores from technically inclined institutions as well as Yale’s statement regarding STEM major persistence. I think it’s not beyond reason to say that this should be explored further even at this wide scope so as to better understand why. (Ithaca’s report did speculate that the more experiential teaching they do is less limiting to those who might have not scored as highly on the math section. Emily Roper-Doten said something similar of Olin of all places - where she previously worked - again due to the style of teaching and learning there. Perhaps that’s a place to start the exploration of why this variation exists).

Viewed this way, the assertion that each school must decide for themselves makes the most sense, as they know best where they reside along this spectrum, and their data will reflect that. (MITChris: “I think everyone college/university should do what is right for them”). To assert that either boundary case should be true for all institutions seems counter to this bulk of opinion/conclusions (conclusions that are based on local and specific research). At this point, it feels less interesting, at least to me, to further try to drive to the boundary cases… too much work, too much vested interest headed in other directions, etc, etc.

So I think this long discussion re: the NYT article is really a tiff over this “scope”. Akil Bello put it nicely. From the IHE article: “Bello said his primary frustration with Leonhardt’s piece has been the tendency to generalize its conclusions and apply them to all of higher education. ‘In reality, places like The New York Times are almost myopically focused on the Ivies and the most highly rejective institutions,’ he said. ‘It’s like they’ve got horse blinders on. But families are reading it and saying, ‘This applies to everyone!’’” If Leonhardt overstepped, it’s by overstating how wide a scope SAT/ACT’s utility had vs cost. But in his defense, he’s didn’t say every school either: “The SAT debate really comes down to dozens of elite colleges, like Harvard, M.I.T., Williams, Carleton, U.C.L.A. and the University of Michigan.” (NYT).

In fact, Lee Coffin also opined along those lines, a fact that the YCBK noted in ep 382: “he [Coffin] speculated that more schools are gonna be going back to test required as a result of the research”

So I wanted to end this (long) post with an account of the famous minutes on Lee Coffin’s podcast episode. Careful reading does evoke a certain wisdom which is enlightening to the situation. And again, Emily does a great job representin’:

LC: Lee Coffin (Dartmouth)
JQ: Jeremiah Quilan (Yale)
ERD: Emily Roper-Doten (Clark)


JQ:
So at Yale, we were looking into this question before the pandemic just to understand how important standardized testing was in predicting how well a student would do at Yale, and it turns out actually that the SAT or the ACT is the single best predictor of a student’s academic performance at Yale. Um, and particularly the math SAT, in persistence in some of our science majors. Um, this is a bit counter to the national research, which suggests that GPA is a bit more predictive than standardized testing. But at an institution at Yale [sic], um, we find that the standardized testing is the single biggest predictor.

LC:
Yeah, I would just add to you that we’re studying the same thing and that’s the emerging storyline here [Dartmouth] as well.

JQ:
And that’s a valu… so that means it’s an incredibly valuable part of our process, um, I… the other thing I’d like to say about this is that we ground everything we do in context.

[…explains context in calibrating test scores and transcripts…]

But, in the admissions committee room we will often pull up the transcript for the 5 person committee to look at and to examine to help us understand the story of a student’s journey, we’ll never look at the testing beyond just the preliminary glance at the start of the application file, because you know I’ve never been in a committee room where someone said oh my god that collection of SAT scores is so compelling I’m wanna vote to admit this student. That’s just not how it works.

LC:
Same, and it’s, it’s… I think people are surprised by that. Emily, you’re starting to laugh.

[laughter]

ERD:
Well, I’d… I’d… you know… if I may, I’d love to say something for the non-hyperselective set , you know on this point because I think for a lot of us the transcript is the part of the profile that is up for the lead actor’s spot, if I harken back to my theatrical roots right? Um, where that is sort of the most prominent, and the testing, if it’s there, plays a supporting role for that. You know, I think, really it is there as an opportunity to… to buoy, but not to sink. At least for many of us. Um and I love that you brought us back to the idea of context and talking a bit more about what context means for the student. I think it’s also important for families to understand that there is institutional context. Right, we’ve talked about, at our individual institutions, how testing plays a role. How rigor plays a role. And so I… it is important to understand that there isn’t… one categorical answer for many of these things and it’s really our responsibility and I… I hope listeners take away from all three of us that we’re all doing the work on our campuses to understand these things, so that our policies are aligned with who we are as an institution, how we teach what we value…

10 Likes

My question for Jeremiah would be why use stem to support a statement that the math SAT is predictive of success. That is like saying Yale discovered that it’s basketball players less than 6 feet tall tend not to be stars on the team.

Test optional creates a lot more work for Jeremiah and his peers at other institutions, yet most have not made the move back to requiring tests. That’s the tell.

3 Likes

The SAT can be a valuable tool for admissions officers. It may also be that they feel in many cases, they can get to the same point with other information.

In that event, both “the SAT is useful” and “we will be test optional/blind” can both be “true.” Both will end up leaving holes in the admissions sieve, so it’s really a matter of which way you want to err.

5 Likes

Yes, it is each college figuring out if the “juice is worth the squeeze”.

3 Likes

I would be surprised if these schools ever go back to test required. I think they like the flexibility TO gives them. In addition, TO policies have (as I understand) attracted more FGLI and URM applicants to these schools and I think they would be very reluctant to lose that.

5 Likes

I agree. They are two sides of the same coin but it seems the nuance in their motivation is wanting the benefits of TO rather than avoiding the negatives of testing. All components used in admission are wealth correlated and could have been viable targets for removal. But TO provides better returns.

To whom?

Doesn’t seem like most colleges believe that. And again, if some schools think it does help them identify students they want, that’s great and their prerogative.

Separately, Stanford announced they will be TO for another year, for the HS Class of 2025.

2 Likes

Not sure I understand your question. I’m just reiterating what Thorsmom66 was saying with the additional comment re: the flavor of the motivation.

Yale, Brown, and Dartmouth are still TO. But given what has been publicized, I’d be more wary about which students I’d advise withholding scores, at least for these three.

2 Likes

I agree. I am not sure if these particular schools would have gone TO if not for the pandemic, but now that it is here I think it would be very difficult to get rid of it. The recent SC decision cemented my view because that made the kind of flexibility offered by TO even more important to those schools.

3 Likes

Test scores clearly have predictive value, but other factors also have predictive value to the extent that in many/most cases test scores are not necessary. Some institutions based on their own data believe that the juice is worth the squeeze, but what is the juice? It may be a data point that has independent value or a data point that is easier to get than putting together a lot of other factors, probably both.

TO is likely to stay not because tests have no relevant value, but because it has proven to attract a wider pool of applicants and given the SCOTUS decision on affirmative action, the more opaque the decision, the more cover they will have. These elite schools have not changed their position that diversity is important. If TO encourages a more diverse applicant pool (probably less expensive than targeted outreach) and allows them more room to use subjectivity in their selection (and apply “context” to those who do not submit scores), I don’t see them going back to test required unless the total number of applications goes out of control.

5 Likes

In going through the process this year, applying widely (excluding west coast) it seems to me that some schools favor high GPAs & mostly disregard test scores, and others favor high SAT/ACT scores over GPAs. And many look at a combination. That said, I do not see a possibility that schools with 60,000 applicants are reading most of those applications. They might read the bubble applications, but I surmise that there is an algo along the lines of GPA+SAT+STATE+HIGH SCHOOL that sorts for them, depending on their above mentioned priorities. First round picks the highest x# from such state/high school, and down the list.

For all those on the bubble, essays, LOR, LOCI, emails etc might get a student into the admit pile.