Percentage of Applicants Accepted by SAT (or ACT) range?

@Chembiodad I certainly accept the possibility that applicant pools change from year to year. Hamilton’s increase may very well reflect an increase in applicant pool quality compared to its peers. I haven’t looked at all the NESCAC schools (D19 isn’t interested in all of them). A total of 15 schools I follow have reported, mostly at or slightly below Hamilton selectivity (a few above as well). Of those, I think Lafayette did better (total jump of 50 points from M+CR to M+EBRW), as did Lawrence (67 points). Williams also had a 30-point jump. Other schools might not have had as large an overall jump, but might have been like Amherst, which jumped 15% in 700+ Math scores and 15% in CR->EBRW scores. A fair reading would say that it’s unlikely that Hamilton’s jump is due entirely to change in scoring, and it’s very possible that their applicant pool is improving faster when compared to many of their peers, but the entire jump is unlikely to be due just to the peer issue.

Again, I can’t emphasize enough that none of this thread is meant in any way to be a slight on (or praise for) Hamilton. It’s merely a thread trying to figure out whether students and families can use the CDS and score band data to help give a slightly better idea of relative level of chances, regardless of whether they’re applying to Hamilton or elsewhere. Hamilton happens to be the first school that has posted both a 2017-2018 CDS and a Class of 2021 score band. (MIT and Princeton will post their 2017-2018 CDS at some point.) I’m hoping to look at Amherst’s 2016 scores next (since we have the Class of 2020 score band for them) not in any attempt to compare Hamilton and Amherst, but to see whether the rough acceptance “bonus” ranges apply to them as well. I’d love for schools with a slightly less competitive selection – think Macalester or Mount Holyoke or Reed or… any place else with a 30-50% acceptance rate – to do this as well.

@homerdog, I was only comparing the CDS 2017-2018 (first link) as compared to CDS 2016-2017 since we don’t have similar information (second link)) from most schools.

In addition to Amherst, Vassar seems to have seen a similar increase in the Bottom-25% and % above 700.

Ok. Still not understanding. The second link still shows the number of kids accepted per SAT band. And it shows the number of those kids who decided to enroll at Hamilton. Since we all know that ED is binding, it makes sense that all of the kids who were accepted and didn’t matriculate applied RD. That’s all I was trying to say. Maybe I’m just trying to make myself feel better since S19 will not use ED when he applies to school and, at least at Hamilton, you can see a decent percentage of kids in the upper band got accepted RD. Am I still not understanding the info in that second link?

@BorgityBorg, small quibble, the 75th percentile for the new SAT is going to be below 1500. Based on the table provided, 24 out 138 matriculated students scored 1500-1600 or 17.4%. In terms of accepted students it was 32.8% (121/369) which may be more relevant in a sense for candidates trying to gauge their competitiveness relative to the applicant pool.

Does anyone know why the CDS does not include a composite for the SAT, but does include one for ACT? I am annoyed by that missing bit of data even though I may be able to find it on some schools’ websites. I’m also unsure how to figure reaches/matches/safeties where there is a disparity in section scores, one high (at or above the 75th percentile) and one low (say, around the 25th).

You know one thing really strikes me about the Amherst CDS regarding testing. The college is in Massachusetts, deep in SAT country. And traditionally more (enrolled) students submit the SAT.

In 2003 85% did SAT, 15% ACT.
https://www.amherst.edu/media/view/7880/original/first_year_admissions.pdf

In 2016 after gradual annual change away from the SAT, it was almost 50/50 but SAT still won out slightly (w/51%).

In 2017, BOOM. 33% SAT and 67% ACT.
https://www.amherst.edu/system/files/C%2520Admission_1.pdf

Now I wonder if this trend will last past the uncertainty of the newness of that SAT redesign…

@BorgityBorg, agree that’s it clear from Hamilton’s total application increase, ED I application increase and median score increases that something is happening. Time will tell.

Given that both the Common Application and ED process have been in play for many years I don’t think students are submitting to an ever increasing # of schools as applications are only up 10%, but I do think students are re-assessing the use of ED.

Our DD’s had friends that profiled more in the middle-50% to bottom 25% at many of these highly selective schools and got Denied, and then went on to get WL’d in RD everywhere except at their more safer schools. I think people are realizing that unless a student is well within the top-25% or has a significant hook it will likely result in a WL or Denial. As a result, I think ED is being used more for what is perceived as a lower reach because a student is in or around the top-25% so it feels realizable, as the alternative outcome of ending up at a safety isn’t very attractive.

Regarding my post #125 I see at Hamilton ACT has also overtaken SAT this year (36% SAT/47% ACT) vs (61% SAT and 8% ACT in 2005).

Brown: 96% SAT/23% ACT in 2004 vs 67% SAT/49% ACT in 2016 with no CDS available for 2017 yet.

Williams also had more ACTs submitted this year, but just by 1%.

Vanderbilt, in more favorable ACT territory, reports 22% SAT and 74% ACT this year, though they were 80% SAT and 37% ACT back in 1999.

Maybe this thread should be discussing ACT scores rather than SAT scores…for certain schools, at least.

Agreed w/ @OHMomof2 – I’ve been discussing SAT because that’s what I (and we as a family) are familiar with. To the extent that ACT scores are sometimes reported as composites, similar to how the relatively few score bands are reported, it may give slightly more accurate (or slightly less inaccurate) results given the individual SAT section weirdness compared to SAT total score bands.

Interesting tidbit, wisteria.

Thing is, many kids don’t use their applications to their advantage, fail to show they do give a hoot about that college.

Homerdog, that’s not proclaiming love, sending multi emails or visiting as much as how you show your match. After all, when a kid does care, he/she does tend to dig a little deeper, get to know more than US News, reputation, that the school has your major, etc.

@lookingforward, absolutely agree. Far too many misunderstand the difference between demonstrating interest and demonstrating fit.

@InfiniteWaves yes, I agree that a student should do their due diligence and truly apply where they can really see themselves thriving. i hope S19 will stand out in his essays and show why he matches a school and why a school matches him. He’s not throwing darts at a dartboard making his list.

Circling back to the Amherst 2016 data (again, only because besides Hamilton’s 2017 data, it’s the only school I’m aware of that has available CDS and score band data from the same year). I am not particularly interested in Amherst’s specific acceptance rates. Heck, because of the way they report it, it is very easy for a student with a specific score to figure out their rough chance of acceptance. I am more interested in whether there is some relationship between the CDS and score bands so that people can look at CDSs for other schools who don’t report score bands and get some sense of their relative safety (or not).

Basic facts from https://www.amherst.edu/system/files/media/Amherst%2520College%2520SSR%2520Class%2520of%25202020.pdf

Acceptance rate: 14% (page 1)
25-75 percentiles (page 2) (enrolled students, per CDS methodology, and these match their 2016-2017 CDS):

  • SAT Math (old) 680-780
  • SAT CR (old) 680-780
  • SAT W (old) 680-780
  • ACT Composite 31-34

Acceptance rates at 75% level:

  • SAT Math (old) [750-800]: 22% (369/1712)
  • SAT CR (old) [750-800]: 25% (350/1401)
  • SAT W (old) [750-800]: 23% (342/1466)
  • ACT Composite [34-36]: 23% (257/1098)
    So… maybe a 1.5-1.75x acceptance bump?

Acceptance rates at 50% level:

  • SAT Math (old) [700-750]: 15% (130/890)
  • SAT CR (old) [700-750]: 16% (159/1015)
  • SAT W (old) [700-750]: 16% (157/989)
  • ACT Composite [30-33]: 15% (244/1594)
    So there’s virtually no bump at being at roughly the median.

Acceptance rates at 25% level:

  • SAT Math (old) [650-690]: 14% (102/753)
  • SAT CR (old) [650-690]: 13% (97/734)
  • SAT W (old) [650-690]: 13% (102/794)
  • ACT Composite [30-33]: 15% (244/1594) [I know, same band as the 50% range! Maybe it’s better to assume someone at the upper end of the range might be, like, 18% and near 31 might be at 12%?]
    There is very little difference, it would appear, between the 50% and 25% acceptance bands. Once you fall below that 25% level, it seems to drop off more, but this is a much flatter acceptance range.

Acceptance rates at 25% and below (matching methodology in Hamilton example):

  • SAT Math (old) [below 700]: 11% (153/1433)
  • SAT CR (old) [below 700]: 9% (143/1619)
  • SAT W (old) [below 700]: 10% (153/1580)
  • ACT Composite [below 30-33]: 6% (50/873) – probably underestimates since some portion of 30-33 range would be in here.

This does not look exactly like Hamilton’s data, for which there was much more of a bonus associated with scores at or above the 75% percentile (and penalties at 25% and below). Not sure if that means there’s a point at which schools whose acceptance rates fall below 15% have such a high percentage of good scores, that there’s not nearly as big of a bump with a super higher score? We’d need to see the MIT and Princeton CDS for 2017 to see if that theory holds water.

Again, this data for schools with acceptance rates higher than Hamilton (and Amherst) would be really helpful.

@BorgityBorg From your research and spreadsheets, can draw any conclusions about new vs old SAT scores? The original concordance table predicted new scores would be much higher than old. Do you see that?

Vassar has good score distribution data https://admissions.vassar.edu/about/statistics/
https://institutionalresearch.vassar.edu/docs/VassarCollegeCDS-2017-18.pdf

^A third of the Vassar applicants submitting SAT scores submitted Old ones. I think it’s interesting that the medians and means for Old and New scores (700-710-ish) are all below the midpoint of the middle 50 range reported in the CDS (720 for both math and EBRW). Perhaps this illustrates how the 2017-18 CDS is less than an accurate reflection due to its forced reporting of Old scores as higher concorded New numbers per the 2016 tables.

@wisteria100 – simply recording the change from 16 Math+CR to 17 Math+EBRW yields an increase of 34 at the 25th percentile, 21 at the 50th, and 8 at the 75th. Not entirely surprising that as you get closer to a 1600 max that the increase diminishes. But if you compare it to the 16 Concorded score to the 17 score, it’s a decline of (37) for the 25th percentile, (33) for the 50th percentile, (28) for the 75th. Scores went up… but not as much as the College Board concordance tables would have predicted.

That’s based on 15 schools I’ve been tracking that have actually reported the 17-18 CDS. (I’ve got another 20 or so who haven’t yet reported.) Primarily selective LACs – the college with the lowest score that I’m tracking is Agnes Scott, who clearly isn’t Caltech, but is no slouch, either. I don’t think a decline in the quality of the applicant pools for the 15 that have reported is an issue in this case.

@evergreen5 – I agree completely, though I think it’s also possible that some schools concorded their 16-17 CDS to the new SAT. Hopefully it’ll mostly work itself out by the time the 18-19 CDS are reported.

@BorgityBorg, That makes sense as Vassar, as a most selective school and one which has done a great job of tracking Old SAT and New SAT results over the last two years, has only shown a small increase in median SAT scores - that assumes the distribution is the same, which I haven’t confirmed.

@BorgityBorg Thanks for that reply! I have heard that College board is not going to re-do their old vs new concordance table as it was based on a sample of test takers and they are not going to recreate another sample group, but it does seem their table could prove to be a bit off. For those applying this fall, the CDS data that will include SAT scores from just the new test, will be very relevant but not available in time, so will just have to monitor school website data and news releases. Interesting though, because I think College Board did readjust percentiles for the SAT, perhaps acknowledging their original concordance could be off at the high end.
I do think students will begin to return to the SAT. Think there was an unknown about it and some thought that the test prep material might not be as relevant or abundant in the early days of the test, but now those issues are no longer.

@wisteria100 High school class of 2019 won’t have CDSs to see, but the College Board is going to release a new SAT/ACT concordance table this summer with just new SAT info. I know it’s a little off-topic since this thread is discussing Old and New SATs bands, but this new SAT/ACT chart will indeed help 2019er’s understand their SAT scores a little better.