This repo contains the code to replicate the Scannel and Kurz Post on why huge swings in app growth might not always be good for Enrollment Managers.
One important note. I have no affiliation with S&K. I simply wanted to build upon a common idea and produce code so others can run and build upon this very important concept in Enrollment Management.
About I have been collecting and analyzing data along the lines of the S&K and post for a while now, but they beat me to the punch and did a great job highlighting an important point.
TL:DR Below I use Prismatic’s API to tag the mission statements of approximately 500 colleges in the U.S. in order to evaluate the “focus” of each, which I define as the topics extracted from the API. In addition, I also consider the competitive nature of various schools, commonly referred to as the “Competitor Set”.
Out of the gate, I considered more than 1,000 schools for this study. In the end, I only kept institutions with clean data, that is, no missing information across all of the data I collected.
Intro The use of graphs to solve business problems is not new, as companies like Amazon, Netflix, and nearly all major social media sites have been doing this for some time. I have been obsessed with graphs for just as long, and after learning as much as I can about analysis of graphs and graph databases, I am finally getting the time to take what I have learned and apply it to real world data problems I face within Enrollment Management.
TLDR; Enrollment Science is my attempt to relate the emerging role of Data Scientists to Enrollment Management Divisions within Higher Education. Let’s get real. Higher Ed is BIG business, and it’s about time that we start to embrace advanced techniques (i.e. the methods industry has been using for many years now) in order to be more efficient. We need to think differently about how we run our organizations, and in my opinion, it starts with the principles of Enrollment Science.
We were talking about admit rates at work, and as luck would have it, I saw the tweet below coming out of the Eduventures conference basically the same day.
More than 70% of American colleges accept everyone who applies. Only 200 colleges accept less than 50% of applicants #eac14 — Karlyn Borysenko (@KarlynMB) June 5, 2014
Because we can download full survey datasets from IPEDS, it’s fairly simple to hack something together fairly quickly to test this ourselves.
As of late, there has been a surge in conversation around the topic of the college-going population here in the United States.
One one hand, we have long talked about the “The Perfect Storm” of demographics. For example, here is a simple Google Search. On the other, the decline in college enrollment, has been connected to changes in the labor market.
In the end, it might be nice to review what data exist and highlight how these flashy headlines could have been predictable well in advance of 2014.
Reaction: Netflix-like Recommendations for College Earlier today I saw the following post:
Could a Netflix-like algorithm be the key to matching students to the right college? http://t.co/6oPgM62zA9 #HigherEdData #collegeaccess — Chronicle Data (@chrondata) January 23, 2014
Before I go deeper into my thoughts on the matter, I want to give a hat-tip to Dan Jarratt for actually deploying his idea. Well done sir.
Dan is on to something and I commend his approach.
Abstract: A simple example of predictive analytics for Enrollment Managers using FREE tools.
TL;DR: Using R we can fit all sorts of complex models in Enrollment Management, quickly, and for no cost. In truth, data modeling can help undercover complex relationships at your school that are not easily visible in our usual tables and charts. However, predictive analytics is not the golden ticket to enrollment success. You will need to understand not only what the model is telling you, but also the risks associated with being incorrect.