Based on pre-TO data only 25% of kids from households at the lowest level of income were taking a standardized test. Undoubtedly, that % is lower today.
1 Like
A quote from a WSJ article on Dartmouth:
âSacerdote said one of the most surprising findings for him was that an applicantâs high-school GPA might not be all that useful a predictor of college success. Students with a perfect 4.0 in high school earned college grades that were just 0.1 point higher than those with a high school GPA of 3.2, according to the study of Ivy-Plus schools.â
This highlights the issue that many of us who question the diminishing value of GPA given rampant grade inflation and believe test scores are a valid additional data point. Out of my public HS class of 230±, there were only 2 students with 4.0âs (no weightings back then). Extra credit and retests were not part of the vocabulary. In my kidsâ public HS class of 400+, there were probably 20 UW 4.0âs. Maybe 5-6 of those kids scored 30 or higher on the ACT.
10 Likes
âWeâre looking for the kids who are excelling in their environment. We know society is unequal,â Beilock said. âKids that are excelling in their environment, we think, are a good bet to excel at Dartmouth and out in the world.â The admissions office will judge an applicantâs environment partly by comparing his or her test score with the score distribution at the applicantâs high schools, Coffin said. In some cases, even an SAT score well below 1,400 can help an application.
This argues against the notion that they are primarily using the SAT to determine future success, since they are also using the test result in context of each individual application and not necessarily applying a universal standard. Maybe they are requiring a minimum floor to accept certain applicants?
There was an interesting post (maybe this thread, maybe another) from someone who did analytical modeling at a state university about how SAT scores aligned with college graduation success rates. This person said:
- SAT scores do align with freshman year success.
- Freshman year success was the greatest indicator of whether or not a student would graduate.
- They then insisted that SAT scores did not offer relevant data on whether or not a student would graduate.
This didnât make sense to me. I hope this Dartmouth study at least makes this person reconsider the âconclusionsâ of their model.
2 Likes
I donât know what study or thread you are referring to, but this Dartmouth study applies to Dartmouth onlyâŠnot to other private schools, definitely not to publics.
1 Like
Are you open to the idea that what Dartmouth found to be true could likely also be true for other public and private universities?
2 Likes
The same thing was said about MIT when they reinstalled tests.
1 Like
Sure, but only with data specific to the school. Note that the data Dartmouth is basing this decision on (at least the data surrounding 1400 test scores) is from when they required tests and also when they could include race in their admissions decisions.
The element of context allows them to consider low-income students coming from underperforming schools, so there is a chance this is a better solution to this issue of diversity than just checking a box on a form.
Yep. And mitchris said this:
â100% and itâs very frustrating to me, personally, that we get used as a club to bludgeon other schools. The MIT education is super weird and specific and essentially begins with a series of high stakes math tests in the GIRs, so of course math tests have predictive utility. I can imagine that for some other schools this may not be true; indeed Iâd be surprised if it werenât at least somewhat different elsewhere!â
and this:
"There are other explanations:
- A school may have done research and found the SAT/ACT doesnât predict for them
- A school may not have done any research at all
- A school may think that it has a different mission than optimizing academic success at their own institution as measured by certain indicators, and that mission could be considered normatively good or bad by others depending on their own goals"
Well, sounds like Dartmouth has done their research.
6 Likes
I think they are simply saying âif you made the most of the resources available to you, we believe you can be successful here.â
3 Likes
Yep â it sounds like they will be judging scores⊠holistically. (in context)
Sure. But it is also possible that Dartmouth defines success and builds a student body differently than a public university. I mean that is clear if you even look at the SES of the student bodies of an average state flagship and the highly rejective, high end private school. There have been a number of other studies that indicate differently and itâs likely those results sit in some other context.
The other thing about a public flagship or a more regional college/university might have is better data on how students out of individual schools likely to apply tend to do in their program. Their missions are different.
2 Likes
It may be more applicable to some kinds of other schools than others. If HS grade inflation means that the most selective colleges are choosing between applicants with 4.0 GPA, then HS GPA becomes a poor predictor for them (the few admits with significantly lower HS GPA are likely so exceptional in other aspects that correlation to college GPA may not say much).
On the other hand, HS GPA may be much more useful to moderately selective schools. But such moderately selective schools are probably seen as beneath students and parents on these forums, so they do not exist for the purpose of forum discussion, even though they are far more numerous than the schools focused on here.
I donât know if GPA or the SAT matter for those schools, since most of them accept the vast majority of applicants.
A year from now, other than the UCs (and other schools bound by law) I expect most highly selective schools will have shifted back to requiring standardized tests. Weâll see.
3 Likes
Visit the âParents of HS Class of XX with 3.0-3.4 GPAâ for that discussion. It is active
1 Like
SAT testing IS worth it for the low income kids who get more financial aid from top privates than from in-state publics, leading to a lower cost of attendance or potentially free.
4 Likes
yes, I understand* but the vast majority of HS counselors in CA are more worried about getting kids to graduate and into a Cal State or UC as opposed to informing them that a full ride from ~30 highly selective schools is a possibility (with slightly better odds than winning the lottery).
*My kid attended one of the highly selectives for less than the out-of-pocket cost to attend our flagship UC at instate rates.
5 Likes
Not disagreeing with your statement, but what exactly is the added value of the test score requirement in assessing success in light of the resources available in comparison to an application without a test score?