Analyzing Retention Data from the Successful Student Perspective

Photo Courtesy: Wikimedia Commons

Photo Courtesy: Wikimedia Commons

Recently we were asked by University Business to respond to a few questions on using data to support student success. They were curious about the approach of looking at the profile of successful students vs. focusing just on at-risk student profiles.  Our senior data analyst, Jon MacMillan, provided these responses:

How important is it for college administrators to be examining data to piece together a profile of successful students, as opposed to focusing mainly on identifying at-risk students?

For a long time higher education has been using data to identify at-risk students based on any number of factors. For most that used to just include things like mid-term gpa or end of term gpa, depending on when you wanted to identify the at-risk students. However, schools have started to investigate other relationships in regards to student success and student completion. While many still focus on creating early alerts for those students at risk, more and more schools are starting to create a successful student profile.

There are a number of benefits of being able to identify those students that are more likely to succeed at your institution. First, you might be able to identify these attributes based on information that you would have at the application stage. That way, you would be able to shape your class not only based on the same desirable traits that have guided your admittance selections previously, but now with the added benefit of knowing what each applicants probability of success is at your institution.

Additionally, instead of just understanding what the key indicators are of at-risk students, if you instead create a profile of successful students you will be able to identify areas for improvement. For instance, while I was working with a community college we found that students that took a certain developmental math course in their first term were more likely to continue on and complete their degree. This information is invaluable to institutions. Often times there are factors that are out of the institution’s control, but being able to identify the markers of a successful student, especially those within their control has a dramatic impact on student success and overall institutional integrity.

What advice can you offer campus leaders looking to analyze their retention data in this more “positive” way?

If you are looking to analyze your retention data in the more positive way you need a tool that allows you to pull in all of the disparate data that you have available. The data preparation is 80% of the workload when it comes to predictive modeling, so when looking for products make sure that you either already have a tool that can do this for you, or the product that you are looking to purchase has these capabilities. Often, our first thought is to go straight to our student information system and use the data available there, but we need to remember that students leave a footprint all across campus and a lot of the time this doesn’t show up in our student database. You need to keep an open mind to all of the different avenues for capturing student information throughout your institution and a tool capable of extracting that data.

One of our customers, Ball State University, was using the information that they had within their Banner Student Information System as the foundation of a student success model, but in addition to that, they started to investigate all the other data elements that they had been tracking. They were looking at campus involvement through sporting events and other organizations, campus conduct violations, as well as when and how often students were eating breakfast. Their hunch was that students that attended breakfast earlier and more often, were students that were more likely to succeed at the institution. Through another database system, they were able to track this card swipe data and create a count of the number of times a student attended breakfast throughout the term and incorporate that into their Banner data. In the end, they were right and students that attended breakfast more often and earlier were in fact more likely to be retained and continue their education.

Schools in the past have focused their student success models on student performance and while this is generally a good indicator of student success we know that this alone does not capture the entire student profile. In the past couple of years there has been tremendous growth when it comes to data collection and we are beginning to see institutions take advantage of that. Now on top of the pre-existing known indicators of student success, schools are starting to look at card swipe data, census data, demographic information, financial aid offering, as well as many other factors. In the end we are left with a much clearer picture of not only who is going to be successful, but why.

Rapid Insight recently sponsored a webinar on using student data and predictive analytics to boost retention. It was presented by Kay Bales, Vice President for Student Affairs and Dean of Students at Ball State University. In case you missed it, you can watch an on-demand version of the webinar at any time. 

Decentralize analytics.
Harness the power of many.

Create and share reports and datasets across the enterprise, and put analytical power in the hands of everyone. Veera creates a truly data-driven culture. Try it for yourself today.



Decentralize analytics. Harness the power of many.