BI Analytics for University Admissions
A good BI data model is like good industrial engineering — form and function blend just right. The form part is how natural it feels to work with. The function part is the range of utility. Like any good tool it should be easy to use, and versatile.
Let’s explore how this might unfold in the context of the university admissions process. The admissions process is a careful balancing act seeking to enroll the right number of the best possible students, while maximizing faculty capacity and facility teaching space. Factors include student academics, diversity, financial aid, acceptance yields, and a host of others; a perfect problem for a strong BI analytics solution.
Ok, so we need to craft a data model that helps to optimize admission planning and forecasting, degree program and faculty scheduling, and the general makeup of the admitted student population. First we turn to the trusty dimensional model as our frame of reference for devising the right kind of model (by “model” I really mean the tool – easy to use, and versatile): and its core principal is to model around the business events that are occurring.
Ah, so you say “we want to optimize the admissions process” and so the business event is “a student is admitted”. True, but only a tiny part of the story and not all that powerful by itself. The key is to think about the process itself: a student applies for admission, they are offered a place of admission, they accept the offer of admission: three business events! And even that isn’t the whole story. We should consider the process before applying and, for that matter, the process after acceptance.
By doing so we ultimately end up with a model that reflects the lifecycle of the student. It goes something like this: it starts with showing interest in a degree program (e.g. downloading a program prospectus), it moves on to complete an admissions application, being offered a place of admission, acceptance of the offer, showing up for enrollment, potential switching of degree programs, academic progress milestones, potential termination or resignation, eventual graduation/alumni and donor. The key business events that map to this process are thus: prospect > applicant > admitted/pre-matric > registered/in-progress > alumni > donor. It is from this vantage point we design our model.
And now you can see the power this brings to admissions planning because all of these events feedback into the prospective students you want to attract and admit. Think about it for a second: this model allows you to uncover if certain kinds of students tend to switch degree programs. Wow, and just as powerful, the model is easy to use and incredibly versatile. It is easy to use because it naturally relates to the way the various administrative processes work in the University. It is versatile because it can support multiple different purposes: for example, I can better plan out my faculty ratios by understanding the historical and probable degree program changes within a certain student profile. Or, to take it even further, I can even characterize my future alumni giving opportunity.
If you want to learn more about this, we will be presenting this concept in detail at the upcoming Higher Ed DW conference in April. If you happen to be attending, please come and say hello.
In a later blog we’ll describe the design of the model, the key data supporting the student lifecycle, and the key dimensions that bring it to life.
Thanks for reading, please send us feedback, comments or other perspectives, we always love to hear from you.