The subject of this project is to predict variables that are significant in a student’ s enrollment choice. Sophisticated statistical tests were used on data collected through Cornell’ s Admission Office for students of the Class of 2009. The information was used to predict whether or not an admitted student would choose to enroll at Cornell given that student’ s characteristics. By omitting insignificant variables from the model, interesting trends were discovered. Many of these trends were intuitional, but a few were surprising. Probit regression analysis strongly suggests that factors such as distance from Cornell, an applicant’ s high school GPA, and religion, among others, play a role in determining if a student will enroll.
Luke Behaunek, ‘06 Belle Plaine, IA
Majors: Mathematics, Economics and Business
Sponsor: Todd Knoop