Someone on CC had previously posted a chart showing calculations (may have been Stanford) on the changing acceptance rates for kids with a 35-36 ACT (or SAT equivalent) and how often those kids were accepted over kids with impressive but lower scores.
Is it worth the time and effort (and, for some people, money) to raise a 32-33 to a 34-36 or to get that SAT from a 1540 to a 1570, etc.? Do you think the kid with the 34-36 is significantly more likely to be accepted to a school if GPA and EC and other things are somewhat similar? Let’s assume we are talking about top 40 colleges, not just HYPSM, etc.
Of course it depends on the school - I’m just interested in general as an observer of people spending alarming amounts of money on test prep.
As a general statement, I would say that if the scores are in the 25th percentile for the college in question, and the applicant is not hooked, then, yes, it is worth taking again to try to achieve a higher score. Of course, for any of these schools, stats are just one part of the equation. So while low stats will hurt, higher stats won’t guarantee admission.
Whether one needs to spend “alarming amounts of money” on test prep is a separate question.
@ucbalumnus , yes, except I don’t like that the chart lists ACT as 30-36 because that doesn’t break it down enough. And while I did say data because I’d like to see it, I’m also interested in opinions. I know someone who spent over $6,000 in test prep. Some families spend more. But I guess those families sometimes justify that the $6k investment can turn into $50k or more in merit aid, or different options for matriculation.
I’d love to know if the 34-36 or 1550+ really does provide a tangible advantage in and of itself if the student has respectable but not “hooked” other credentials.
Situations where there is a clearly defined threshold score (e.g. the threshold for National Merit on the 11th grade PSAT, or the SAT or ACT threshold for automatic-for-stats scholarships at some colleges) make it relatively easy to determine whether retrying is worth doing. But these situations do not apply to super-selective colleges with holistic admissions processes, so many parents and students are making their own guesses as to how valuable an extra point on a retry could be.
I agree with @ucbalumnus that it’s probably more important to raise that ACT to a 34 or higher for Merit reasons than for admissions in many cases.
Also this is pure speculation/opinion but schools that publish the breakdown like Stanford and Brown don’t put as much weight on the tippy top scores as much as schools that don’t publish that kind of breakdown. Those schools don’t want to confirm that a 36 or 1600 is a big advantage in admissions.
For ACT, 32-33 are 97th-98th percentile and 34-35 are 99th percentile, with 36 having percentile “100.”
For SAT, the 99th percentile is 1480-1520 and for everything above, the percentile is designated 99+.
I have a hard time believing that, inside the 99th percentile on the SAT, a few questions at often 10 points apiece leads to a significant difference in quality of applicants. (Obviously, I’m not an adcom.) However, looking at middle 50 ranges, there does seem to be a difference between 32 (97th percentile) and 34 (99th).
As an aside, I’m getting annoyed when selective schools (like the Stanford link) don’t report data for the SAT composite. It makes it difficult to evaluate where a student might stand if one section score is significantly higher than the other.
Adding: SAT EBRW 750-760 is 99th pecentile, 770+ is 99+. SAT math 770-790 is 99th percentile, 800 is 99+.
^^^^ I actually interpret the Brown data differently. While it is true that 72% of applicants with a 36 were rejected, if we look at the admit rate between the 36 vs 33-35 pool, it is 28% vs 11%, 2.5x higher. If we slide down to 29-32 range, it drops further to 7% (33-35 applicants having a 1.5x higher admissions rate). While there is likely correlation that applicants with higher test scores probably also have better grades/rigor of courses/LORs, essays, I can’t see that much of drop-off in those qualities between “36” applicants vs “33-35” applicants. So I do see a big difference between having test scores well above the historical median vs at or below, and it is not a case of all applicants being “equal” after a threshold and being judged/compared at that point purely on the subjectives.
@BKSquared , I shared the same interpretation of Brown’s figures. I think it’s safe to extrapolate that while many of those kids with a 36 had equally-impressive other factors, I’m inclined to think that not all of them were distinguishable from other strong applicants.
Moreover, perception plays a large part and no top school is ignorant to the fact that higher-average scores is associated with a better reputation at least in some respects (and also in rankings).
@BKSquared “I can’t see that much of drop-off in those qualities between “36” applicants vs “33-35” applicants…” Based on the studies I have seen it is very likely 36 kids have higher rate of national awards and other distinguishing ECs than 33-35 kids. There are two common misconceptions I think—one is that test preparation makes a big difference, in fact its only 0.8 or so on average; and the other is that among top 1%ers there is no difference in accomplishments and the fact is top 0.25% has far more accomplishments than top 0.75%-1%. If AOs treat 33 and 36 the same (as most of them claim) they would have seen significant differences in awards/accomplishment between the two groups.
Actually, I don’t necessarily like lumping 33 and 35 together either. I see the 35 scorers in a closer range with the 36 scorers and would like to see the breakdown in outcomes by number versus in a span of three or more. Some of those 35 kids missed a 36 by one question. The same can’t be said for those in the 33 range.
Therefore, when I see that Brown’s data shows an 11% acceptance rate for those in the 33-35 range versus 28% for those with a 36, I can’t help but think the 35 scorers are somewhere higher than 11%.