Predictive Modeling and Student Success

There are only so many resources an institution can devote to improving student success. Thus, it would make sense to spend those resources where they are needed the most if we are to maximize our student success results. 

This is often what is lost in "predictive analytics." Data are not a strategy, they are a tool that help us achieve our strategic goals. Moreover, with all the data in the world and the most sophisticated predictive models, if we don't change what we do to support students, those efforts will be wasted.

Graph picture_edited.jpg

Thus, DIA's work in predictive modeling certainly emphasizes data analyses, but more importantly data use and impact. We not only help identify predictors of student success, but place those within your institutional context, understanding key populations of interest, potential action steps to improve student success, and the organizational change necessary to enact predictive modeling data. 

DIA's predictive modeling work takes place in several key stages:

  • Reviewing existing data to determine which factors are available and which might be considered in future modeling efforts;

  • Identifying key local factors, such as populations of interest, high impact practices, and essential student outcomes;

  • Multi-modal reporting that goes beyond a technical report to provide a clear understanding of the process, lessons learned, and action items to improve student success;

  • Training and support for a variety of staff on how to engage with students based on their likelihood for success, strengths, and challenges.​


Many predictive analytics ​efforts help predict success, but they do little to drive institutional action that can improve success. Our work in this area is the embodiment of our credo: data to information, information to action.

For more information about our work in Predictive Modeling, simply fill out the contact information below, entering "Predictive Modeling" in the subject line.