SAT concordance table - compare old and new SAT scores

It is hard to compare ACT and SAT from individual data points. My D2 got 33/1540. There are always students perform better in one test or the other particularly when they only take the test once or twice. The percentile charts of the two test are more meaningful. However, the old SAT score concordance table using scores from students taking both tests should be even more meaningful and hopefully there will be one for the new SAT score soon

More new Class of 2022 data

FSU http://www.fsunews.com/story/news/2018/02/11/fsu-receives-record-breaking-number-applications/326977002/ (I think this is for all accepted students, including RD)
SAT 1290 - 1400
ACT 28 - 32

UVA accepted EA http://www.cavalierdaily.com/article/2018/01/uva-accepts-27-8-percent-of-early-action-applicants-for-class-of-2022
SAT 1390 - 1520
ACT 32 - 35

BC accepted EA http://bcheights.com/2018/01/17/3170-admitted-early-action-class-2022/
mean SAT 1453
mean ACT 33

Georgia Tech accepted EA http://www.news.gatech.edu/2018/01/12/tech-admits-4677-students-through-early-action
SAT 1390 - 1530
ACT 32 - 35

Georgetown accepted EA http://www.thehoya.com/class-2022-early-action-applications-hit-record-high/ Difficult to compare because they are not reporting an SAT composite.
SAT “verbal” 710-770, SAT math 720-790
ACT 32 - 35

@evergreen5 Thanks for posting more new info. How are you finding these numbers as schools make them public? Do you check certain schools every once in a while and just google “name of school and EA SAT scores”?

Feeling better and better about S19’s 1540.

@homerdog I occasionally google Class of 2022 admissions under News and look at the articles. Sometimes the stats are in the articles themselves, but other times they’re contained in an attached graphic, in which case the stats aren’t searchable.

Bear in mind that these are for accepted students. With the large increases in apps everywhere, it’ll be interesting to see the enrolled student data for Class of 2022 that’ll be released in the fall.

@homerdog I am doing exactly what you’re suggesting @evergreen5 might have done, except that I have a bookmarked list of schools’ institutiona data/CDS pages that I check every weekday. (It’s how I channel any college-related nervousness so I don’t share it with D19. I love data.)

I’ve tracked down 24 schools’ scores – there are more, no doubt, but these are the ones that might conceivably be in D19’s interest area. There are another dozen or so on my bookmarked list still waiting to be posted. These are generally schools in the top 50 or so, but no Ivies, no Stanford, no Chicago.

For the 24 schools that have posted their 17-18 CDS, the median “slip” from their 16-17 CDS concorded to their 17-18 CDS varies by percentile. At the 25th percentile, the “slip” is 10 points – meaning, for example, a school’s 16-17 CDS might have translated to a 1320 SAT at the 25th percentile, but their 17-18 CDS reported a 1310 at the 25th percentile. The “slip” at the 75th percentile is greater – 20 points. (The implied slip at the 50th percentile, where I’m just assuming halfway between the 25th and 75th percentile unless specified by the college, is 11 points.)

Having said that, there’s tremendous variability, to the point that suggests some schools have messed up their concordance calculations because it’s unlikely a school jumped or fell so dramatically in just one year. For example, Beloit (https://www.beloit.edu/irap/collegedata/cds/) would have been expected to have a 75th percentile score of 1435 based on their component scores in 16-17, but only reported a 75th percentile of 1310, or a decrease in concorded scores of 125! Meanwhile, Williams’ 75th percentile scores jumped from 1540 to 1570. But they were one of only 2 schools (Lawrence being the other) whose concorded scores jumped at the 75th percentile.

While the median is useful, I will note that it’d be nice to get more of the schools in slots 10-30 like Bowdoin, Haverford, and Grinnell. Those are schools for which the concordance matters for your S19, @homerdog, because if you take their concorded scores and knock 'em down 10-20 points, the 75th percentile is just under 1540. But if they beat the trend, then they’re above…

@BorgityBorg I do it for the exact same reason :slight_smile: but I don’t check often. I just wish next year’s CDS were available earlier, before apps are due. Keep up the good work!

@evergreen5 I choose to control what I can control, and that’s checking a list of links during a lunchbreak. :wink:

I just checked my Google doc’s version history, and I had a dozen of the 17-18 CDS scores – half of what I have now – by Dec. 21 last year. So while it’s unlikely that most schools nationally will have updated by then (and it’ll be random as to whether it specifically helps your own kid’s application process), I think it’ll be enough to determine general trends for the 18-19 CDS – the trends I indicated above were all there in roughly the same magnitude two months ago.

I’m confused. I thought these new EA stats are not concoursed but actual new SAT scores. Is that not correct?

Sorry if I wasn’t clear, @homerdog – the 17-18 CDS should reflect new SAT scores – either actual new SAT scores or old SAT scores (for those kids who submitted pre-2016 SAT results) concorded to new SAT scores. What I’ve done with my analysis is take the 16-17 CDS SAT scores – which were supposed to be reported as old SAT scores (including concording any new SATs to the old score) – and concord those to the new scores. So I was trying to project what the 17-18 SAT scores would be based on the 16-17 SAT scores, and then comparing those projections to the actual 17-18 scores.

It actually wouldn’t surprise me if the 18-19 scores drop slightly again as the “inflated” (when concorded) pre-16 SAT scores drop entirely (or virtually entirely) out. It’s also not clear whether the clear jump in preference for ACT over SAT during this transition time will continue or if it’ll shift back more in favor of the SAT, and if so, how that affects the CDS.

@homerdog the ones I posted above for Class of 2022 are virtually all New SAT scores, albeit mostly for early apps.

The 17-18 CDS that @BorgityBorg is working with gives data for Class of 2021, with a mix of Old and New scores, and Old first concorded to New using the College Board’s tables and then mixed in with New. From what I understand, the proportion of Old scores for last admissions season may vary quite a bit from school to school.

We won’t have CDS for Class of 2022 (with virtually all New scores) until late fall at the earliest, with most coming after apps are due, so we’ll be relying on the data usually posted as a Class Profile on each college’s website. There can be some inconsistency in what colleges decide to post, e.g. admitted vs enrolled, SAT section scores vs composite, whereas the CDS is always for enrolled and always for section scores.)

For the life of me I cannot figure out why ACT gets a composite spot on the CDS but not SAT.

So class profiles are usually updated on schools’ websites before their CDS is available?

Yes, if only because there’s a lot more detail in the CDS? (That’s a presumption on my part based on the small amounts of data in the class profile and the large amount in the CDS.)

That’s setting aside the issue of if the profile reflects accepted as opposed to enrolled students (as in the CDS).

Right. Plus who knows if the enrolled profile looks better or worse than the accepted profile. For all of our purposes, at least from an admissions standpoint, accepted student numbers seem to be the most important. Although, it’s important to see which kids actually attend as well.

Ah, that’s something else to occupy my data-loving brain! Trying to figure out whether the enrolled profile does look better or worse than the accepted profile. I would suspect that it varies from school to school based perhaps on its perceived prestige. This is probably something that doesn’t affect the tippy-tops with 50%+ yield, but it wouldn’t surprise me if there’s a general trend where some of the highest scorers decide to attend a different, more competitive school. That to me seems how colleges are now using ED – lock in the high scorers and folks meeting a variety of demographic needs you know want to attend the school, and then use RD to fill in the gaps however you want (test scores, other ECs, etc.).

[/ thread hijack]

I also love data. Some schools don’t have the actual CDS on their sites but instead display the admissions data in a different format. I feel like they are trying to hide it. Is there a place where the CDS data is published other than the college site? US news and other sites get the data from the CDS and use it to input into their websites but don’t actually link to the source doc (which they should). Just wondering where u could find actual CDS files for prior years.

@suzyQ7 Try https://nces.ed.gov/collegenavigator/ (Disclaimer, I can’t confirm for certain that this is CDS data, but that is my understanding. I wonder when the 17-18 data will end up here.)

Colleges that publish the CDS on their websites do allow you to choose prior years from a list.

Note that CDS has **enrolled **freshmen stat so it is only submitted after the school year started. Admitted student profile is available much earlier and the school pretty much know the incoming class after May 1, so the admission stat can be posted much earlier.

Thanks KS @evergreen5 . I think it is based on CDS but not presented in the format of the actual CDS and lots of info is missing. Wish they included the actual CDS.

Spoke with someone affiliated with admissions at a competitive school and was told to disregard the concordance table entirely. They said for their admission guidelines a 1400 SAT “concords” with a 34 ACT. And based on EA results from this school that seems in line with the 2022 results