Do you mean highly rejective schools? I lean that way too for that set of schools, but hope the schools that choose that policy make it really clear, rather than the gray area we are in now…‘X school really likes test scores’, ’ X school expects them from affluent kids’, etc.
2 Likes
We heard similar stories. S24 took the SAT in August and we had to drive an hour away to get a seat - there were no testing sites in our area. Our HS didn’t even host a March sitting last year as it has done every other year (even during Covid).
1 Like
Maybe/at the margin, but I don’t think D announcing they are now test required changed the testing plans of many students.
I agree, just noting that the issue with other schools having little time to now announce doesn’t really change things. It’s like if Sally were now to decide she had to test for (additional schools that announced test-required) Sally would already be screwed.
1 Like
Yes. I think it goes against Harvard, Princeton, Yale’s (etc…) mission to do anything that will seemingly reduce access.
We should point out that Princeton and Dartmouth (I haven’t checked others) never moved testing from the “very important” column in the CDS. And Princeton has stuck with language that reflect that. I was actually surprised they didn’t bring it back/extended it for multiple years, even two years ago, but understand the optics issue. From the website:
“Though standardized tests results will not be required for fall entry 2023, 2024 and 2025, we still value these results and will evaluate them within the context of our holistic review.”
“We consider all of these measures within the context of each applicant’s school and situation. Our most promising candidates tend to earn strong grades and have comparatively high scores on standardized tests.”
This is completely different from say, UVA. I have seen Dean J state multiple times something like “We don’t need a test. That is just a number. It doesn’t tell us anything about you.”
3 Likes
August is particularly tough as there are not as many sites. I think its because a lot of districts aren’t in session yet.
2 Likes
By replacing the actual data points with averages, and not even adding error bars to the averages, they create an illusion that weak relationship is far stronger than it actually is. It is a deceptive visual, and no academic journal would ever permit somebody to present a graph like that. But they are not only presenting averages, they are also binning the scores.
The actual data has far more points along both the X and Y axes, and the “data points” on th egraphs not only represent average mean GPA for that SAT score, but also average SAT score for that average mean GPA. So they are the averages for SAT scores versus averages for mean GPAs.
They are essentially, eliminating most of the variance, in order to create the impression of a far stronger relationship than there actually is.
No peer reviewed journal would accept an article with that sort of graph, specifically because the graph is misleading.
3 Likes
So, my characterization of your comment is correct?
Unconditional, blanket statements about something are often risky to make, and can often be wrong. Because all you need to disprove them is a single counterexample.
There are countless academic articles that present tables and figures containing median income of a demographic, average life span of a species, median commute time to work, average velocity in N trials, average entropy of a system — all without including information about the standard deviation, skewness, etc., of the underlying distribution.
An average is what it is. One can question what it doesn’t capture, but to say that no academic journals would ever accept manuscripts reporting only averages is going overboard.
5 Likes
Do you feel Dartmouth is being deliberately dishonest to carry out an agenda, are they just wildly incompetent, or is it just some level of recklessness that is in between the two?
1 Like
[
huh? It’s one of oldest engineering schools in US (though admittedly not one of the biggest !)
3 Likes
Sure – apparently I had the 3/2 part backwards. Not Dartmouth sending kids to finish Eng degrees elsewhere, but the other way around.
3 Likes
At a public university I know, the importance of tests in Admissions varies depending on majors.
For most majors, curriculum rigor with A/A-/B+ is basically the only thing that matters. The SAT verbal will help with placement but if no score is provided they know how to evaluate the curriculum. Evidence of rigor is sufficient - honors, AP, IB, DE, with scores or predicted scores or A grades all work.
However for math-heavy majors they really want to see a math score - SAT or ACT would be the most common though 5s in AP Calc and AP Physics 1 (for instance) also work. They’ve noticed that for these majors in particular, test optional led to a -.2 GPA variance in pre-req grades. TO with better entering grades did as well as test submitted+ slightly lowerGPA.
This was taken into account by Admissions.
In short, students, to have the best chance, should either submit a Math SAT 630 and a 3.65+ GPA, OR submit a 3.8 GPA without APs OR submit 2-3 STEM AP 5s and A/A- grades. But if students had a 3.5-3.6 with no AP score and also went TO, their odds of academic success had been shown as below average and therefore they’d be on the bubble/WL.
(This is not public but if you know where to dig it’s been published then discussed officially.)
5 Likes
This makes a lot of sense, and I’m glad there are data that support it.
The fly in the ointment here is the lower ceiling in both gpa and testing of late (i.e. grade inflation and lower testing difficulty). They dull the tools with which admissions must execute their tasks. As usual, the unintended consequences of good intentions sometimes hurts as well as helps.
I wish we lived in a clearer world.
A number of schools have not yet announced their plans for 2025: UPenn, Duke, Brown, Carnegie Mellon, and Emory at least. (In November I predicted that some schools that hadn’t announced yet would not remain TO for 2025: I’m 1/3 for the ones I guessed so far (Cornell and Stanford are, Dartmouth is not)).
Cornell has different test policies for different schools. A summary is below. The different schools also emphasize and require different non-score criteria in admission decisions. For example engineering requires taking calculus and having a math teacher LOR. It also recommends taking CS. In contrast Hotel requires a resume and emphasizes having resume experience in the hospitality industry. I get the impression that they aren’t going to mandate that all schools use the same testing policy, which makes sense given the varying admission criteria.
- Test Blind – Architecture, Business, Life Sciences, and Hotel
- Test Optional – Arts & Sciences, Ecology, Engineering, Labor Relations, and Public Policy
1 Like
Interesting article/info from Adam Ingersoll at Compass Education. Includes a graphic with current testing policy status at selective schools.
3 Likes
Most interesting lines…
SAT scores are powerful inputs for our algorithms predicting an applicant’s likelihood to enroll
SAT scores are powerful inputs for our algorithms predicting an applicant’s willingness to pay
5 Likes
I was list to the Yale admissions podcast. They were telling the applications increased from 26k to over 50 k over last 7-8 years. So they started a new method of screening applications. Senior AO would do a quick review and reject some applications- who had no chance of getting into Yale based on some criteria. They would then pass on the rest to the regional AO. They said this saved a lot of time. Also said that a lot of increase is amongst not so competitive apps. Some of the criteria used by senior AO were English proficiency, academics, citizenship , maturity and ability to thrive at Yale.
1 Like
maturity and ability to thrive at Yale.
How does one determine this with a quick review?