Anonymous wrote:

Anonymous wrote:

Anonymous wrote:The number from a 2022 Common App report:

76,747 students applied to college with ACT/SAT scores >1500 (99%)

Wow. I assume the number is so high due to superscoring? This is an eye-opening figure for sure.

Yes, because of superscoring and many coached kids take the test numerous times to get to 1500, superscored over two tests or from taking it several times until they reach that score on the final take.

I'm in higher ed and have access to data from the Common App and Slate (for research purposes) and people would be surprised how big the pool of 1500/34+ pool is, especially in the last 10-15 years.

Forgot to add that it was close to 90K in 2024!

Anonymous wrote:

Anonymous wrote:The number from a 2022 Common App report:

76,747 students applied to college with ACT/SAT scores >1500 (99%)

Wow. I assume the number is so high due to superscoring? This is an eye-opening figure for sure.

Yes, because of superscoring and many coached kids take the test numerous times to get to 1500, superscored over two tests or from taking it several times until they reach that score on the final take.

I'm in higher ed and have access to data from the Common App and Slate (for research purposes) and people would be surprised how big the pool of 1500/34+ pool is, especially in the last 10-15 years.

Anonymous wrote:The number from a 2022 Common App report:

76,747 students applied to college with ACT/SAT scores >1500 (99%)

Wow. I assume the number is so high due to superscoring? This is an eye-opening figure for sure.

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:I recommend that you don't use superscoring in this analysis. Superscoring seems like a public relations gesture to stressed out students. Rather than something that really impacts a lot of candidates to get them admitted. And you don't have any data on what improvements are typically seen with superscoring. Any assumptions might be quite off.

?? A 35 superscore is the same as a 35 in one setting. Test scores check a box not get you admitted. It is possible you could create a good impression by a one and done 35 but you would create a better impression by a 36 superscored. All colleges are looking for is the number. It does impact a lot of candidates.

No, it's not. It implies a different (lower) "true ability".

No it doesn’t. The SAT/ACT don’t test aptitude at all. They will both tell you so. If you can’t assume tests across different test dates are within a very narrow range of difficulty then that renders these tests practically useless as a comparitive metric.

They are g loaded tests.

The tests are equated between test dates. https://www.collegevine.com/faq/17181/how-does-the-sat-curve-work

The tests are the single most important predictor of performance at highly selective colleges.

You do know that this isn't true based on copious past research which shows GPA to be slightly better than test scores as a predictor. I believe that because of grade inflation and GPA compression at the top this will reverse as newer research studies are completed but to date the research refutes your assertion.

Most recent research shows test scores are more predictive than GPA; almost certainly because of grade inflation. The best predictor is test scores, grades and rigor assessed together.

I agree that this is where it is heading but we need to see more study. I only know of one single small example but logically more research will confirm the change.

Anonymous wrote:A very high percentage of students scoring above 1500 and 34 are going to state flagships. The cost of private universities has really changed the dynamics of acceptances and yields. Often these kids are getting substantial merit, and that really changes everything when people are looking at a $400,000 bill per kid for private colleges.

This. My 1570 score kid isn’t even considering any T20 schools. It is far from universal that every top scoring kid is shot gunning top schools, or applying to any of those schools.

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:I recommend that you don't use superscoring in this analysis. Superscoring seems like a public relations gesture to stressed out students. Rather than something that really impacts a lot of candidates to get them admitted. And you don't have any data on what improvements are typically seen with superscoring. Any assumptions might be quite off.

?? A 35 superscore is the same as a 35 in one setting. Test scores check a box not get you admitted. It is possible you could create a good impression by a one and done 35 but you would create a better impression by a 36 superscored. All colleges are looking for is the number. It does impact a lot of candidates.

No, it's not. It implies a different (lower) "true ability".

No it doesn’t. The SAT/ACT don’t test aptitude at all. They will both tell you so. If you can’t assume tests across different test dates are within a very narrow range of difficulty then that renders these tests practically useless as a comparitive metric.

They are g loaded tests.

The tests are equated between test dates. https://www.collegevine.com/faq/17181/how-does-the-sat-curve-work

The tests are the single most important predictor of performance at highly selective colleges.

You do know that this isn't true based on copious past research which shows GPA to be slightly better than test scores as a predictor. I believe that because of grade inflation and GPA compression at the top this will reverse as newer research studies are completed but to date the research refutes your assertion.

Most recent research shows test scores are more predictive than GPA; almost certainly because of grade inflation. The best predictor is test scores, grades and rigor assessed together.

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:I recommend that you don't use superscoring in this analysis. Superscoring seems like a public relations gesture to stressed out students. Rather than something that really impacts a lot of candidates to get them admitted. And you don't have any data on what improvements are typically seen with superscoring. Any assumptions might be quite off.

?? A 35 superscore is the same as a 35 in one setting. Test scores check a box not get you admitted. It is possible you could create a good impression by a one and done 35 but you would create a better impression by a 36 superscored. All colleges are looking for is the number. It does impact a lot of candidates.

No, it's not. It implies a different (lower) "true ability".

No it doesn’t. The SAT/ACT don’t test aptitude at all. They will both tell you so. If you can’t assume tests across different test dates are within a very narrow range of difficulty then that renders these tests practically useless as a comparitive metric.

They are g loaded tests.

The tests are equated between test dates. https://www.collegevine.com/faq/17181/how-does-the-sat-curve-work

The tests are the single most important predictor of performance at highly selective colleges.

You do know that this isn't true based on copious past research which shows GPA to be slightly better than test scores as a predictor. I believe that because of grade inflation and GPA compression at the top this will reverse as newer research studies are completed but to date the research refutes your assertion.

A very high percentage of students scoring above 1500 and 34 are going to state flagships. The cost of private universities has really changed the dynamics of acceptances and yields. Often these kids are getting substantial merit, and that really changes everything when people are looking at a $400,000 bill per kid for private colleges.

It gets even wackier when you take into account the upper tier huge public colleges like UCLA, Berkeley (all the UC schools), Florida, Texas. They get a big share of 1500+ SAT scorers. You even add in schools like Purdue, Illinois, Rutgers, Michigan and realize they get quite a few high SAT scorers (maybe not as many as a percentage of the student population), but a decent amount of the raw total.

Anonymous wrote:Because I'm bored... if there are roughly 195,000 undergraduates students in the top 20... and roughly 48,750-ish incoming freshman... and only 12,000-ish 99th percentile test scorers... doesn't that mean only 25% of these students are in the top percentile?

That means don't worry if your kid doesn't score in the 99th percentile. There is plenty of room for them. Go ahead and apply and use your scores.

Sorry no, supper scoring will cover that not to mention schools recruited athletes and other priority needs. So that is why really good students get shut out. So not 49,000 seats available. Maybe 30K seats available and super scoring will raise the number to 50K students looking to fill 30K seats.

* total T20 seats available/ unhooked >1500 applicants
( replace above)

There are roughly 2 million SAT takers and 1.4 million ACT takers. A proportion take both.

Assuming there are 3 million test takers. 1% yields 30,000 who are in the 99th percentile.
--------------------------------------
meant to say 3 mil is def double counted and not unique applicants

Anonymous wrote:There are roughly 2 million SAT takers and 1.4 million ACT takers. A proportion take both.

Assuming there are 3 million test takers. 1% yields 30,000 who are in the 99th percentile.

With superscoring, that number at least doubles, so one is looking at 60,000 students scoring in the 99th percentile.

The number from a 2022 Common App report:

76,747 students applied to college with ACT/SAT scores >1500 (99%
---------------------------------------------------


2021-22:
7.6 mil applications from 1.2 mil applicants (roughly 6 applications per applicant on avg)

applicant SAT/ACT >1500 ~ 645000 (~6% of applicants)

so # of T20 seats (unhooked)~ 0.5* 2000*20 ~20000 seats. ( 50% reduced to account for 20% INt, 30% ur/donor/hooked)

so % of unhooked / total T20 seats available ~20000/ 450000 (no of applicants reduced by 30% to account for Int/urm/donor/hooked but not proportionally assuming success rate of ~70% for urm/hooked/donor, iNt rates reduced proportionally)

so % of unhooked/total T20 seats available =5%

so give and take that % is in the range of 5-20% (say). So Score>1500 alone does not cut it for T20 and we know it///

now apply the constraint: full pay (can be afforded by 90% net worth families)

Voila, the fog lifts !!!!

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:I recommend that you don't use superscoring in this analysis. Superscoring seems like a public relations gesture to stressed out students. Rather than something that really impacts a lot of candidates to get them admitted. And you don't have any data on what improvements are typically seen with superscoring. Any assumptions might be quite off.

?? A 35 superscore is the same as a 35 in one setting. Test scores check a box not get you admitted. It is possible you could create a good impression by a one and done 35 but you would create a better impression by a 36 superscored. All colleges are looking for is the number. It does impact a lot of candidates.

No, it's not. It implies a different (lower) "true ability".

No it doesn’t. The SAT/ACT don’t test aptitude at all. They will both tell you so. If you can’t assume tests across different test dates are within a very narrow range of difficulty then that renders these tests practically useless as a comparitive metric.

They are g loaded tests.

The tests are equated between test dates. https://www.collegevine.com/faq/17181/how-does-the-sat-curve-work

The tests are the single most important predictor of performance at highly selective colleges.

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:I recommend that you don't use superscoring in this analysis. Superscoring seems like a public relations gesture to stressed out students. Rather than something that really impacts a lot of candidates to get them admitted. And you don't have any data on what improvements are typically seen with superscoring. Any assumptions might be quite off.

?? A 35 superscore is the same as a 35 in one setting. Test scores check a box not get you admitted. It is possible you could create a good impression by a one and done 35 but you would create a better impression by a 36 superscored. All colleges are looking for is the number. It does impact a lot of candidates.

No, it's not. It implies a different (lower) "true ability".

No it doesn’t. The SAT/ACT don’t test aptitude at all. They will both tell you so. If you can’t assume tests across different test dates are within a very narrow range of difficulty then that renders these tests practically useless as a comparitive metric.

They can tell you what they want but the whole idea of testing and all the statistical analyses around it is based upon a premise that there exists an underlying ability and that tests are sampling from that ability (the ability is not necessarily IQ). Your best guess at the underlying ability is the average of your scores. If you are only looking at the highest score, you are overestimating ability.