College admission stats vs. commitment stats

<p>I am new to this forum (wish I had found it much earlier!). But it seems like there is a discrepancy between class profile admission stats that college publish, and actual commitment stats - am I right? </p>

<p>After the admission process closes the newpapers and colleges typically publish the stats for the "new freshman class profile" admitted - (e.g. "the 2012 freshman class admitted to College X has an average SAT score of x, uweighted GPA of 3.9, etc. etc.). But then colleges start pulling off their waitlist, which means the those admitted who comprised that average didn't commit. Presumably the "profile" stats go down too? Maybe not enough to matter. Am just curious.</p>

<p>The web figures are usually students who actually enrolled as freshmen. Some schools will also list stat categories and the number admitted from those tiers. Not many. In any early media article, they could state the results for admit letters- but a chunk of that total pool often chooses to enroll elsewhere. Many colleges pull from waitlists per their own policies. Some go top down by stats, some pick on other bases. You’re right that some of this can be confusing. Most people, I believe, just look at the stats for enrolled kids.</p>

<p>The Common Data Set (example here <a href=“http://www.swarthmore.edu/Documents/administration/ir/cds2011.pdf[/url]”>http://www.swarthmore.edu/Documents/administration/ir/cds2011.pdf&lt;/a&gt; ) has the info. Section C1 shows how many were admitted and how many enrolled. The stats shown below that, section C9, are for those who enrolled.</p>

<p>Thanks!!! That helps!</p>

<p>Basically, all the stats you see are post-wait list activity.</p>

<p>Brown gives pretty good information about accepted students and enrolled students separately: [Admission</a> Facts | Undergraduate Admission](<a href=“Undergraduate Admission | Brown University”>Undergraduate Admission | Brown University)</p>

<p>You can see that the differences aren’t that dramatic, but there’s a noticeable shift in the distribution of test scores. So, for instance, about 32% of the enrolled class had a CR score 750 or higher, but 40% of those admitted had 750 or higher. Basically, when you look at the 75%-25% range for enrolled students, you should add 10-20 points to each to get an idea of the range for admitted students.</p>

<p>The other thing to understand with very selective colleges, especially the small LACs (like Amherst or Williams), but relevant up through places like Harvard or Stanford, is that the bottom quartile of test scores almost entirely consists of students who were recruited for something – usually sports, but it can be special music talents, or being a famous actor, or having a unique, compelling story. Applicants with those scores and no such hooks have virtually no chance of admission, not even the low chance that the numbers imply. Which means that while the Brown common data set will tell you that 25% of enrolled students have a CR score of 740 or higher, really the top 25% of accepted students excluding recruits has a CR score of 790 or higher.</p>

<p>The stats often tell you a lot about very little. My alma mater Williams, for example, accepts more than 40% of its class via Early Decision, and admits almost 45% of all ED applicants. Yield for those is virtually 100%. Then they admit only around 10% of all applicants in the RD round, and yield is very low (20% max). But you’d never be able to find anything of this reflected clearly in either the admitted or the enrolled stats.</p>

<p>mini, you are not quite being fair to Williams. First of all, it provides 10 years worth of ED data on its website. And that shows that, while the ED percentage seems to have crept up during the period from around 35% to around 40%, Williams has only gone over 40% twice in the past 10 years, and one of those times was 40.8%. Second, it’s also perfectly easy to estimate Williams’ RD yield from the information it provides (or from its common data set). The RD yield at Williams is generally 30-33%, give or take, which is actually pretty darn good. Only a handful of colleges have RD yields that are significantly higher, and those are generally the ones that get referred to here with a single initial, or are members of “the Ivies”. </p>

<p>I suspect Williams’ RD yield reflects the fact that it is admitting really strong students who wind up with lots of great options (and many of whom did not have Williams as their natural first choice, or else they would have applied ED). Getting about a third of them reflects how attractive Williams really is.</p>

<p>How does the waitlist fit into all of this? Many schools are hedging bets with the waitlist, I understand.</p>

<p>Schools don’t generally report their numbers (and don’t generate their common data sets) until the fall, so they are all post-wait list activity. What happens with wait lists is that often schools check with candidates before actually extending an admissions offer, so that the number of accepted waitlist students is only slightly higher than the number of enrolled waitlist students. So the waitlist tends to make the school’s yield seem a little higher.</p>

<p>I’ll use approximate Williams numbers to illustrate: Williams wants a class of about 540, and admits about 1,200 students to get it. But 230 of those students are admitted ED, so by the time RD rolls around they are only trying to fill 310 slots. So maybe they accept 930 kids, and waitlist 700-800 more. If they get a 33.3% yield, their class will be full, but their yield is usually a little below that. Let’s say they get a 30% yield on their RD acceptances, or 279 kids. They need another 31 warm bodies – and probably they look at their presumptive class to see whether for institutional purposes they need those bodies to be male, female, white, “of color”, domestic, foreign, football players, classics scholars, etc. Then, by judicious use of the phone to gauge interest, they only have to issue, say, 35 waitlist acceptances to get 31 more students. </p>

<p>At the end of the day, they have admitted 1,195 kids to fill the target class. Super job, right on the nose! Of course, if they really had had to fill the class RD, they would have had to admit not an additional 35 students, but an additional 104 students, or 1,264. Which is still pretty close to the expectation, but 5% too high, so the pats on the Director of Admission’s back will be less fervent.</p>

<p>Most of the very selective colleges accept very few from their waitlists – probably not enough numerically to have a significant statistical impact. </p>

<p>For those that are taking a larger percentage of students it may be a mistake to assume that waitlisted students have lower stats. I think colleges sometimes waitlist high stat students who they think are unlikely to enroll (that is, the college thinks they are being used as a safety). </p>

<p>Also, stats aren’t the whole story – it may be that most waitlisted students have stats that are consistent with the rest of the admitted students, but simply lacked other stand out qualities. It’s not so much that they were lacking but that they didn’t have the oomph factor that made the cut. I mean, the difference between an admit and a waitlist in a competitive atmosphere could be as simple as a well-written but prosaic essay, and LOR’s that were lacking in specific detail.</p>

<p>If you subtract both the applications and admissions from Williams ED (with virtual 100% yield) from their total applications, admits, and enrollment, you quickly find out that the yield isn’t anywhere close to 30-33%. (And the acceptance rate is about 10%). But it doesn’t matter. The point is that the overall stats may tell you very, very little.</p>

<p>For example, of the 239 admits ED (making up 43% of the class), there were only 14 international admits. Since international admits can cost the college a lot of money if they need aid, it is useful to admit those ED who can pay the entire cost (hence freeing up funds for the RD round.) I expect you’d find that with other parts of the ED process as well, as they seek to lock down full-pays, athletes, and the much needed bassoon-player (and legacies with a demonstrated record of family contributions to the college.) There’s absolutely nothing wrong with this - in fact, I’d run things exactly the same way (especially as not everyone is going to be happy in Williamstown). </p>

<p>Anyhow, I’m not sure the profile admits get at the real issue of who applies, who gets admitted, and who attends, and I think most prestige colleges have this down to a real science. They know precisely what their financial aid budget is in their a-hem needblind process, they have the list of who the athletic departments want, they have the dollars contributed figures for the legacies. They know what they are getting. I wouldn’t expect any top-flight institution wouldn’t. But I’m not sure how you get there from the published stats.</p>

<p>Since Brown is one of few schools that break out applicants by rank/scores and lets you see admit rates, everyone should read it and weep. When you see 19% of vals got admitted and 17% of those with SAT M 800, it brings home how misleading too much focus on stats alone can be. And, how important your full package really is. </p>

<p>The value in placing your kid in the top freshman stats quartile is iffy. Using Brown as an example, the top 40% of current freshmen had M 750+. Your kid may have 780. You think, “Super.” But, that M score tier was accepted at 13%.</p>

<p>Granted,the chart fails becase you can’t see a breakdown that combines CR and M, plus rank. But, it’s sobering.</p>

<p>The main purpose of the waitlist seems to be to try to insure that the precise desired number of seats are filled in the fall, where summer melt is still a confounding factor (more-attractive schools are simultaneously processing THEIR waitlists). At non-wealthy need-aware schools it can also help with the qualification vs. budget balance, taking the minimum number of full-pay applicants needed to maintain solvency. Small schools may also use the waitlist to finely craft the incoming class. What else?</p>