What about the high school students taking the ACT/SAT test over 5 weeks - one hour section per week, in a private room with a private proctor? This appears to be a popular trend amongst private school students. I personally know of at least a dozen or so… not too mention inflated test scores.
The scoring is not this simplistic. Your raw score of number of questions correct/incorrect is converted to a score on a 200 to 800 scale in a non-linear way that depends on the specific version of the SAT you took. For example, two questions wrong on the harder Jan 2018 SAT would give you an 800 math, while 2 questions wrong on easier June 2018 SAT would give you a 750 math. On the SAT practice #1, 1/4 math questions correct corresponds to a 385 (14.5 questions correct). And 1/4 correct on reading + 1/4 on language corresponds to a 350, for a combined score of 735. I suspect some of this lower end score conversion is to make the scores more user friendly, rather than just normalization. It sounds better to get a score of 735/1600 than it does to get a 25%.
Colleges by SAT Score Tiers, 1960
645-675
Amherst
Carleton
Columbia
Harvard
Haverford
Princeton
Reed
Rice
Swarthmore
Williams
Yale
615-645
Brandeis
Brown
Chicago
Cornell
Dartmouth
Hamilton
Johns Hopkins
Lehigh
Oberlin
Rochester
Stanford
580-615
Antioch
Bowdoin
Duke
Kenyon
Michigan
Middlebury
Northwestern
Pennsylvania
Iowa
Tufts
Union
UC–Berkeley
Sewanee
550-580
Colgate
Denison
Grinnell
Knox
Lawrence
Muhlenberg
Occidental
UColorado
490-550
Beloit
NYU
Pittsburgh
Southern Methodist
Syracuse
Virginia
Vanderbilt
Well, at least I’m starting to feel like the apps I put in for myself in the 80’s weren’t totally out of line, even if I didn’t get into the most selective schools I applied to, at least I was somewhat competitive.
I was beginning to feel like maybe I was delusional back then when I see my kid’s friends and teammates and how brutal the admissions cycles can be. We’re feeling really lucky that it is all falling into place for my kid.
What I’m finding interesting about all of this is that there has to be a trickle down effect - so schools further down on the competitiveness list must be getting more applications and in theory, should be getting more competitive too. But how far down does that go? Is there a point where kids and parents are saying “no thanks, it’s not worth it”?
Note that the article from 1960 says the following. If only ~20% of admission was based on SAT scores (I have no idea if the article is accurate for college admission in the 1960s), then it implies that many applicants with scores above the listed means were rejected.
The mean scores in the Life magazine article do not adjust for recentering and SAT test changes. For example, a summary of how the mean SAT scores at Yale is at https://oir.yale.edu/sites/default/files/w032_fresh_sats.pdf and below. I’ve converted to current SAT using concordance tables. The 2018 EBRW mean SAT appears to be nearly identical to the 1976 verbal after concordance adjustment. However, the mean math score has increased.
Yale Mean SAT Scores
1976 – Mean = 670/690 → 740/720 (2017)
1980 – Mean = 670/690 → 740/720 (2017)
1985 – Mean = 670/690 → 740/720 (2017)
1990 – Mean = 670/710 → 740/740 (2017)
1995 – Mean = 670/720 → 740/750 (2017)
<— SAT is Recentered →
2000 – Mean = 730/720 > 740/750 (2017)
2005 – Mean = 750/740 → 760/760 (2017)
← SAT is Redesigned →
2010 – Mean = 750/750/760 → 770/770 (2017)
2015 – Mean = 760/750/750 → 770/770 (2017)
← SAT is Redesigned →
2018 – Mean* = 745/760 → 745/760 (2017)
No direct concordance due to lack of writing section.
*Estimate based on 25th and 75th percentiles
I don’t think looking at SAT scores like this is a good measure of how competitive things are, considering how many other factors go in the decision besides scores at typical highly selective private colleges. It’s quite possible for a college to become notably more competitive while mean test scores have little change or even decrease.
While the higher rated schools are becoming much more competitive, many LACs that US News ranks in the #60 to #120 range appear to be easier admits than they were years back. Some are not filling their classes.
I think the trickle down effect reaches state flagships like Ohio State, and then after that, the schools are starving for applicants.
@privatebanker So, let’s make a very simplistic assumption that everyone takes the SAT three times, and their score increases 50 points each time they take it. The third time you take the test, you are competing against other test-takers who have not yet reached their “potential.” 1/3 of the test takers are 100 points below their ultimate score, and 1/3 of the test takers are 50 points below their ultimate score. Let’s say, also, that once you hit the threshold of the top 1%, you do not take the test again. In this overly simplistic example, I think roughly three percent of the test takers will eventually be able to claim a score in the top 1%, even without superscoring. Obviously, it’s unrealistic, because not everyone will increase their scores by that much from their first to last test, but on the other hand many of these kids will take courses or study between their first and third attempts. Anyway, I think the number of kids who can eventually claim a top 1% score is higher than people think. I’d like to see numbers on this–you’d have to measure results by student ID number, not by the total number of test-takers, to eliminate those multiple attempts by the same individuals.
Regardless, my point is that it isn’t as rare as it should be for a student to get a score of 1500. Now, to get that on the first try, without super scoring–THAT is rare®.
Who knows; maybe my logic is wrong. But that’s the way I see it, anyway. 
I’m only going off the actual statistics published by the college board. The real numbers and statistics.
We can of course and create hypos but no idea if they are meaningful. But you could be correct and perhaps 50k maybe 60k are in the 1500 pool,assuming you are right - out of 2.3mm test takers.
That’s still statistically incredibly rare.
And raising 50 points from 1450 to 1500 is actually extremely difficult compared to lower bands.
I think we use terms like easy and not that hard in cc, it simply is not reflected by the real data. And undermines and cheapens the achievement a bit. Not saying it’s like landing on the moon.
But the actual hard working, college prep student isn’t even in the same stratosphere. 1500 is an outlier and unimaginable to most kids. It just doesn’t seem that way here.
This was our first time through the college admissions process and our son’s experience mirrors many of the points mentioned above. He did use and SAT tutor, took the test 3 times fall-spring junior year and scored 1470, 1500 & then 1520 super-scored to 1530. The weighted GPA after the fall semester was 4.46.
It was a jarring but not completely unexpected to get rejected from Georgia Tech and Notre Dame and then wait listed by UNC. It just goes to show that near perfect scores do not get you into hyper competitive schools. There are a large numbers of applicants with similar scores.
On the bright side, he is accepted to UVA (in state), BC and WPI. Waiting to hear from Lehigh and Tufts.
@cameo43
For a number of years I have tried to stress the importance of the CDS as a research tool. This is particularly true when looking at that tier of Universities which have been historically used as “elite” backups. We are talking here about the apparent increase in applicant levels, but applicants are still functioning on two year old data.
The same problem arises as applicants repeat unsubstantiated opinions about the employment opportunities in a given field. Data is out there, but the horse has to be led to the water.
The “surprises” show up, in large part, because applicants are not using available data. Like a large part of society’s decisions, they are based on unsubstantiated beliefs or on old history.
The foremost example of this problem is climate change. “You cannot fool Mother Nature” was a popular TV advertisement decades ago. The truth comes back and hits one hard.
We need to talk less and study more!
I agree that the Common Data Set can be very useful. It was particularly helpful to me in determining the average amount of merit aid and percentage of students receiving merit aid for students not receiving need-based aid. (Some math was required but the information is in there).
What I cannot find a source for is, for any particular college, the acceptance rate of students in a particular SAT or ACT range, toward the high end of the range. For example, what percentage of applicants with an SAT in the 1500-1600 range is accepted? That would be useful, if anyone knows of such a source.
I did notice that Stanford reports this type of information for broad ranges, but even their information isn’t as granular as one might want. Stanford’s data was helpful, though, in that it had a separate “range” of exactly 800, reporting around 8% of applicants with a perfect math score are accepted and a similar percentage with a perfect ERW score are accepted. Although we don’t really know what this means for the full combined score, it certainly suggests a low admit rate for students having even the best SAT scores (which I supposed we all knew anyway).
If 1560 isn’t feeling “ridiculously great” in this house right now.
That depends on the college and how attractive its options are. And marketing a college has become more active. I also suspect more kids are willing to fly long distances.
Standardized tests scores are widely seen as “the straw that broke the camels back” in college/University decisions.
Keep in mind that the foretasting value of higher education academic success for individual students is NOT well predicted in narrow bands of scores. If one looks at a randomly selected band of 700 MSAT scores and compare them to a randomly selected band of 750 MSAT selected scores, there is very little predictive value and the difference diminishes over time Yet we dwell on it!
Standardized test scores are not IQ test.
This bothers me because we have become fixated on these scores. Both the Universities and the students need to focus more on their very real talent, interests and motivation as a whole. Your talent and the talent of your classroom peers is not strictly defined by perfect standardized test scores. It is more complicated than that.
If one wants to study a math dependent STEM major there are demonstrable relationships, even with Biology and social science majors, as they are now often utilizing a bags of tools from mathematics. If you have a randomly selected group of 600 MSAT scores and you compare their AVERAGE early college performance with 700 MSAT scores, you will find a significant difference, BUT you may still find some big surprises.
I wish we had a math SAT score for Einstein to illustrate this case.
And here we are fixating on the top 2% of scores??
If students/parents would liberate themselves from the marketing term 'top colleges/universities" and focus more on who and where they are, some parents would save a lot of money and the educational system would be optimized.
This does not mean that the lowest cost university is always to best fit. Real employment opportunities also play a role and they are not limited to the “top” colleges.
Do your homework! Get motivated to the right places for reasons which optimize your results.
@Trixy34 I’m sorry it hasn’t been what you’re looking for so far. Hang in there.
Statistical rarity of a score and college outcomes are not what we were discussing.
1560 is one piece of the academic quality ranking which in turn is just one component of several areas.
We say this all the time and no one seems to believe the posters. Once you are deemed qualified they laser in on the rest of the application. And adcoms regularly reveal at the 34 and above level and high gpa and most rigorous level- they move on. Recs essays ecs. You then layer on non controllables. Income level, legacy, athletics, location and ethnicity. And then institutional needs.
There are simply not a lot of 1560s out there. It’s just a numerical fact. And it is a ridiculously great score. It’s just not all that matters.
@trinity2016 I’d like to echo your point that things are getting harder, especially for those applying Regular Decision. Even the Chance Me threads on CC are hopelessly out of date, at least for the non-hooked kids. My DS19 applied with a 3.94/4.71 GPA at a very competitive (median SAT 1490) IB magnet school, fluent in a second language from an immersion program, co-captain of his one major EC, really great recommendations, and a non-superscored 1600 SAT. No prep, no paid counselor. Decided not to apply to any Ivies, Stanford, MIT, etc. just small LAC’s. Got into his favorite (Carleton) but waitlisted at Bowdoin, Swathmore and Middlebury, rejected at Williams. We figured places like Williams and Swat were a lottery, but Williams admitted/aitlisted about 2400 out of 9700, so his application didn’t even make the top 25% there. I was also surprised at Midd and Bowdoin - I doubt Middlebury turned down many profiles like his 5 years ago. The results of the process were great (he got into a great school which is one of his favorites), but it is still surprising. There just aren’t many spaces left for non-hooked kids at the top schools, and seemingly fewer every year. For the few that do get in, their life has to be dedicated to high-powered EC’s and finding a way to stand out in a crowd.
The lesson for our family going forward for the next 3 kids is “Forget about the top schools and just find a good landing place at the publics, lower ranked privates, etc… The life (and sleep) you have to give up to get into one of these programs ain’t worth it”.
“There just aren’t many spaces left for non-hooked kids at the top schools, and seemingly fewer every year.”
That’s my takeaway as well. At top schools - not just Ivies - any school with an admission rate below 20%, RD is a sucker’s bet for the unhooked no matter what kind of stats you have. When the time comes for my DD, it’s going to be ED at a school with an overall admit rate no lower than 25%. We will also have a realistic ED2 school lined up if needed. RD only as a last resort. I wish it weren’t the case, but that’s the way you have to play the game now.
Perhaps even more so at small schools, where the number of athletes to field the usual set of teams is a larger percentage of students than at large schools.
“Stanford’s data was helpful, though, in that it had a separate “range” of exactly 800, reporting around 8% of applicants with a perfect math score are accepted and a similar percentage with a perfect ERW score are accepted.”
Which just demonstrates that a) the SAT is not a high ceiling test and b) Stanford (like other tippy top schools) doesn’t care about your score if it’s “good enough”.
Compare that with the Oxford math admissions test, where a top score does actually make a difference (and at the highest levels, which is approaching IMO qualifier standards, is effectively a guarantee of admission):
https://www.maths.ox.ac.uk/system/files/attachments/Mathsgroup%20feedback%202018_0.pdf (See the chart at the end and note that the candidates taking this test would almost all be self-selecting enough to be in the top 1-2% of math SAT scorers if they were in the US)