Retention Modeling Cohort (Series)

About: The Retention Modeling Cohort is a three-part series that is designed to provide users with practical guidance and time to build and apply models for student retention. Each session builds upon the previous session.

Session 1

In the first convening of the Retention Modeling Cohort, attendees will begin with raw data and leave with a predictive model. The focus of our analysis will be on application characteristics which will have two-fold benefits. First, this means that the data requirements are modest. You can bring more, but you don’t need to worry if you can’t wrangle a large dataset. Secondly, it ensures that your model will be applicable to students from day one! Attendees will receive a heavily supported data preparation process and a user-friendly walkthrough of predictive modeling.

Due to the ambitious goal of leaving the session with a predictive model, there will be required pre-work to gather data before the first meeting. We will provide a worksheet detailing which student characteristics are most valuable when predicting the outcome of retention and participants will be asked to gather as much data in the fields as they can. While participants may feel intimidated by this request, please know that even the most basic data about student demographics can teach us about student outcomes.

Session Recording

Supplementary Videos

The following videos are supplementary materials that were recorded outside of the event.

Session 2

In our second session, participants will score incoming students and build another model. Getting an early source of insight on which students might be at the highest risk of attrition means that your support staff has the opportunity to truly improve student outcomes. Believe it or not, scoring students will be easier than you might imagine so we won’t stop there! In this session, we’ll also discuss a few successful strategies for using the scores and collectively brainstorm a few more. This session will also require some pre-work to ensure that we have historic “first semester” outcomes available. A predictive model founded on incoming characteristics will be impressive but incorporating institutional experiences like extracurriculars and academic outcomes will enhance your model even further.

Session Recording

Session 3

The third session is the capstone for the cohort, but hopefully the beginning of your independent work in predictive modeling for student success. Like the second session, participants will begin by applying the model they built off of “first semester” data and refine their understanding of students’ challenges. Then, we’ll explore the variety of alternate student success outcomes you can predict using these same methods in the future. As we collectively complete our work in this Retention Modeling Cohort, Strategic Leaders will take the opportunity to establish individual check-ins with attendees.
Attendees can expect to see a second round of scores for their students and explore new ways of predicting student success, leaving the series with an improved grasp of predictive modeling and a network of support for all their future analyses.

Session Recording

Download the Session Materials