Use Financial Aid Modeling to Optimize Expenditures. Jon MacMillan, Senior Analyst James Cousins, Sr. Statistical Analyst

Transcription

1 Use Financial Aid Modeling to Optimize Expenditures Jon MacMillan, Senior Analyst James Cousins, Sr. Statistical Analyst

2 Itinerary Who We Are Benefits/Challenges of FA Modeling Different Modeling Techniques Financial Aid Modeling Walkthrough

3 EASY AFFORDABLE POWERFUL ANALYTICS Date Presenter1 Presenter2

4

5 KEY DIFFERENTIATORS Affordable Quick to learn, easy to use Transparency Free and unlimited training and support, we have your back from day one

6 Use Financial Aid Modeling to Optimize Expenditures Jon MacMillan, Senior Analyst James Cousins, Sr. Statistical Analyst

7 Benefits of Financial Aid Modeling Allocate money to the students most likely to enroll Meet more need Increase enrollment yield Allocate money to the students most likely to be successful

8 The Problems Inherent in FA Modeling #1 Financial aid, generally, has the inverse relationship of what we might expect

9 The Problems Inherent in FA Modeling #2 Financial aid is ever changing, which makes getting consistent data challenging

10 The Problems Inherent in FA Modeling #3 Often those performing the analysis are unfamiliar with the packaging process

11 Some Other Inherent Problems Tuition is often increasing General changes in the demographics and makeup of applicant pool Packages are changed during the application process Different/new awards offered year to year Requires being able to access archived data which is often not available Often requires more than just 3 years of data There are many special circumstances that need to be addressed (athletes, etc.) Varying reasons why students might receive more or less aid Clear identification of when the offers were actually made

12 The Problems Inherent in FA Modeling #1 #2 #3 More money = lower enrollment likelihood Awards and amounts constantly changing Financial aid expert needed

13 How Do We Solve These Problems?

14 The Problems Inherent in FA Modeling #1 #2 #3 More money = lower enrollment likelihood We will walk through some examples Awards and amounts constantly changing Look at FA as a percentage of tuition/need Financial aid expert needed Assemble a team and involve them early and often

15 Predictive Modeling Techniques Binning Like Applicants Problem: Students that are offered more money are often the most desirable students, not only to you but also your competitors Solution: Bin students into groups based on similar qualities and then analyze the impact additional money has on enrollment

16 Predictive Modeling Techniques Binning Like Applicants HS GPA HS GPA <= 2.5 HS GPA HS GPA HS GPA 3.5+

17 Predictive Modeling Techniques Offers Based on Enrollment Likelihood Problem: An underperforming discretionary grant designed to address student need and improve enrollment Solution: Within the qualifying pool, identify likely-toyield students, and offer the grant to them

18 Predictive Modeling Techniques Offers Based on Enrollment Likelihood Random Assignment 2 out of 5 offers yielded (40%) Most likely Least likely Ultimately Enrolled Did not Enroll Informed Assignment 4 out of 5 offers yielded (80%)

19 Predictive Modeling Techniques Offers Based on Expected Response Problem: When you have discretionary funds, should you cover all need for some or some need for all? Solution: Identify where additional aid has the largest impact, and offer more to those who respond

20 Enrollment Likelihood Predictive Modeling Techniques Offers Based on Expected Response A difference of 1,000$ may not produce an identical difference in enrollment likelihood B A 1,000$ 1,000$ Financial Aid Offer Amount

21 Predictive Modeling Techniques Binning like applicants Offers based on enrollment likelihood Offers based on expected response

22 Software Demonstration

23 Thank You! For more information visit