6. MODELS ARE MATHEMATICAL EQUATIONS
Y= + 1X1 + 2X2 + 3X3 + ... nXn
Yis the variable to be predicted
X’s represent variables on your students or prospective students
’sare coefficients that are statistically estimated
7. How does a model actually work?
=SLOPE of the line
11. Some examples of models across industries
Direct Marketing:
What is each customer’s probability of purchasing?
How much will each prospect purchase?
Healthcare:
What is each patient’s probability of readmitting within 30 days of discharge?
What is each patient’s risk of Sepsis?
How long will a patient be in the hospital?
Financial:
Which customers will default on their loan?
Insurance:
What is each person’s risk of an accident?
What is each customer’s probability of churning?
13. Prospect/Inquiry Modeling
What is each prospect’s (or inquiry’s) probability of applying?
How can these models be used?
• Buying optimal search names
• Strategically coordinate recruiting efforts by estimating expected
yield
• Prioritize staff and resources
• Mail vs. call vs. email
14. Applicant Enrollment Modeling
• What is each applicant’s probability of enrolling?
How can these models be used?
• Prioritize resources based on each applicant’s enrollment probability
• Help shape your incoming class
• Forecast enrollment and financial outlay based on admit pool
15. Student Retention Modeling
• Which students are at risk of attrition?
How can these models be used?
• Put programs in place to focus resources on at-risk students
• Increase retention and likelihood of student success
16. Alumni Donor Modeling
• Who is going to donate to the annual fund?
How can these models be used?
• Prioritize Advancement efforts based on predicted giving scores.