Big Data

Big Data 

Aaron Fulk
Associate Director of College Counseling
Collegiate School
Big data. We hear about this new phenomenon constantly and how it will change a variety of industries. Personally, I find this phenomenon frustrating on our side of the desk. Despite more information available to us and our students, data does not make this process any less anxiety-inducing or more predictable. No admission dean has shared with me the predictive yield of segmented populations despite my tempting offer of free hugs. Weird. 

With that said, this frustration around data is also one of the few reasons college counselors won't be replaced by robots in the next decade. So we've got that going for us . . . which is nice. With all that in mind, here are a few trends and highlights I've noticed from the most recent admissions cycle.

I feel obligated to point out the obvious about ACCIS: we all work at private institutions with predominantly affluent students. As a result, I am focusing on some of the most selective colleges in the nation, but these colleges and trends are not indicative of higher education at large. Also despite wearing glasses, I am neither an economist nor a mathematician. These are simply personal theories and interpretations of data available to the public. 

In other words, the colleges that are, as the kids say these days, on fleek:

George Washington (22.5% increase in applications, admit rate decrease of 6.5%)
Georgia Tech (12% increase in applications, admit rate decrease of 7%)
Wesleyan (22% increase in applications, admit rate decrease of 5%)

At many of the most selective small, liberal arts colleges, application numbers remained fixed or declined slightly. This trend could signal that these colleges have stopped recruiting uninterested (ghost) prospects as they rely heavily on interest and early applications to narrow yield. Some of these colleges include Amherst, Bowdoin, Carleton, Hamilton, Middlebury, and Pomona. Hopefully, this plateau in application numbers indicate some semblance of predictability in the future . . . or maybe this was simply an odd year.

Given the trend in the last point, I find myself thinking about application increases like annual inflation. As in, if a college does NOT have a 2-3% increase, then they had a "down" year. Obviously, this reaction is crazy, but as you scan down the column of stats looking at application increases from the prior year,the overwhelming majority of colleges experienced jumps of over 2.5%. All of this is to say that I have to remind myself that if a college's application numbers remain fixed -- especially when many of these colleges had record years recently -- or even decline, this is often beneficial to applicants, admissions staff, and (selfishly) us. It will be interesting to see how application numbers continue to rise as projected high school graduates should increase until 2024.

If you take a quick look and compare “regular decision only” admit rates to early decision admit rates, the contrast is stark. This one data point illuminates what I would argue keeps most admissions deans awake in April and May: yield. I found the recent article in The Chronicle, "Record-Breaking Numbers of Applicants? Don't Gloat," compelling. While I would advocate we use yield rates over admit rates when sharing data with our constituents, most of us give board reports in June where yield is still unknown.

I did include 2015 yield rates in the spreadsheets on the ACCIS members-only website as well as predicted yield rates. You should note that these predicted yield rates assume class size remains the same from the prior year and does NOT take into account summer melt. I don’t how helpful these yield rates will be, but again I include them as almost a symbolic gesture that these numbers are more important than application numbers. 

Lastly, HUGE thanks to Steve Frappier, Rob O’Rourke, and Gregg Murray for helping to find these numbers.

Share this post: