This site uses different types of cookies, including analytics and functional cookies (its own and from other sites). To change your cookie settings or find out more, click here. If you continue browsing our website, you accept these cookies.
According to the National Center for Education Statistics, the 6-year graduation rate at public institutions is 61% and the rate at private, non-profit institutions is 67%. And, the national average first-to-second year retention rate among first-time undergraduate students is 81% (NCES). It is particularly important for tuition-dependent institutions to enroll and retain new students while effectively using their limited financial aid dollars to help students be successful. Using predictive modeling and Alteryx Designer, HAI Analytics has helped many institutions predict future student retention and test the effect of increasing the financial aid awarded to students to improve retention.
Describe the business challenge or problem you needed to solve
With increasing costs, changes in financial circumstances, and other outside factors, many students are unable to remain at their enrolling institution. The development and deployment of statistical models in Alteryx Designer can help institutions predict future retention of students and the associated revenues. It can also be used to test the effect of increases in financial aid to students, forecasting headcount, class composition, and net tuition revenue. Using the prep, blend, predictive modeling, and reporting features in Alteryx Designer, higher education institutions are able to set realistic budgets and make strategic decisions regarding the use of financial aid.
The retention of students directly impacts the financial health of an institution, particularly those that are tuition dependent. On the other hand, access to an affordable education is vital for student success. By developing a data-informed approach, schools can find ways to optimize financial aid in a way that is mutually beneficial to students and the institution.
HAI Analytics provides predictive modeling solutions to colleges and business looking to make strategic, data-informed decisions. We have spent our careers in the higher ed space helping schools address their biggest challenges in real time. For many institutions, optimizing financial aid to hit enrollment and revenue goals is their biggest obstacle. By utilizing predictive modeling, schools can accurately predict the number of returning students, as well as test how different interventions will impact future outcomes.
Describe your working solution
To build an effective model for retention, institutions can use the data that already resides in their admissions and financial aid systems. This includes various data points pulled from the application itself (academic credentials, program of interest, state residency, etc.) as well as the amount of financial aid provided. We also can use third-party data from sources like the U.S. Census to create a more robust dataset.
Additionally, many institutions also have engagement data related to the student experience that can be used to predict retention. Because these data sometimes reside in different systems, institutions often extract the relevant data into flat files, which can be easily merged together in Designer.
In June of 2020, we presented a webinar with Alteryx on this very issue. We created a workflow on anonymized data that deployed a developed model. In the workflow, we scored every student with their probability for retention (and attrition) and used that to forecast the expected retention before students returned to campus. Using the developed models, we tested two different strategic adjustments to see how offering additional aid would improve the retention of students and the impact those adjustments would have on class composition, expenditures, and revenues.
We first applied a broad increase in financial aid and compared this to the forecasted retention without any intervention. This provided a way to improve first-to-second year retention rate by nearly a percentage point. We then strategically used the model to identify the 20% of the pool most vulnerable of dropping out, based on where each student stood on a multivariate set of factors in the model. We tested a second, larger adjustment that focused additional aid on these most vulnerable students. In doing so, we identified a way to increase financial aid to improve retention while seeing a gain in the net tuition revenue.
We turned this workflow into an analytic application so that the user is able to test applying different aid increase amounts without having to change the configuration of the tools in the workflow. The user-determined increase in aid is then applied to the pool before running the cases through the model. The report is then dynamically updated to reflect the new adjustment outcomes in the same PDF report.
Describe the benefits you have achieved With Alteryx, higher education institutions are able to easily access and use their data to make informed decisions. Designer offers an easy-to-use code free environment and the ability to turn any workflow into an analytic application. For those institutions needing outside modeling help, we are able to set up customized workflows that provide complete transparency and flexibility.