Academic journal article Academy of Educational Leadership Journal

A Simple Model for Estimating Enrollment Yield from a List of Freshman Prospects

Academic journal article Academy of Educational Leadership Journal

A Simple Model for Estimating Enrollment Yield from a List of Freshman Prospects

Article excerpt


This study develops a simple binary logistic regression model for estimating total enrollment yield and enrollment probability for individual accepted applicants using cross-sectional data from a private, liberal arts university. The model uses three predictor variables: discount amount offered, ACT score and state residence, to predict the enrollment decisions of accepted applicants with slightly over 78% accuracy.


An on-going challenge facing many college administrators is predicting freshman enrollment from year-to-year. Institutions that are highly dependent on tuition dollars, in particular, are concerned about their ability to accurately predict enrollment in order to facilitate the budgeting process. A significant body of research examining the issue of year-over-year college persistence (e.g., St. John, Andrieu & Oesher, 1994; Cabrera, Nora & Castañeda, 1992) indicates that many institutions are reasonably proficient at estimating their retention rate for returning students. However, estimating the matriculation rate for accepted applicants often proves more difficult. Previous research in this area has produced inconsistent results, perhaps due to the idiosyncratic nature and circumstances of the institutions studied. Thus, the purpose of this study is to develop a simple statistical model that can be used to both estimate enrollment yield and predict the likelihood that an accepted applicant will choose to matriculate.


A variety of studies has focused on factors that influence college enrollment and enrollment persistency. In general, these studies have identified two categories of variables believed to significantly impact enrollment decisions. The first category includes academic, demographic and institutional variables such as age, ethnicity, high school experience, gradepoint average, parent's education level and scores on standardized test such as the American College Test (ACT) or Scholastic Aptitude Test (SAT), while the second includes economic and financial variables such as tuition costs, income and various forms of student aid (Braunstein, McGrath & Pescatrice, 1999). These studies have shown that financial aid positively impacts enrollment decisions and that type of financial aid provided (i.e., loans versus grants) interacts with demographic factors such as income and ethnicity to differentially influence the enrollment decision. For example, low-income students have been shown to be more responsive to grants while middle-income students tend to be responsive to both loans and grants (St. John, 1990a, 1992, 1993, 1994).

Because it is dichotomous, numerous studies have employed logistic regression as the primary analytical tool for predicting the enrollment decision of prospective students (e.g., Ledesma, 2009; Goenner and Pauls, 2006; DesJardins, 2002; Bruggink and Gambhir, 1996). Variables that have been used as predictors include a variety of geographic, demographic and social background variables (i.e., race, family income, etc.), academic performance and achievement variables (i.e., ACT and SAT scores, high school GPA, etc.), variables related to student recruitment (i.e. contact with the student prior to application, where the applicant first learned about the school, whether or not the applicant made a campus visit, etc.) and financial aid variables (i.e. discount offered, Pell grant amount, etc.) as well as interactions among these variables.

Braunstein, McGrath & Pescatrice (1999) found that family income, receipt of financial aid, type of financial aid package (i.e., grant, loan, or work-study) and SAT score influence the enrollment decision, and that wealthy applicants and those with the highest SAT scores are less likely to enroll irrespective of financial aid offers. Ledesma (2009) developed a model that predicted enrollment with 64% accuracy and found that high school GPA was negatively associated with the decision of applicants to enroll to a small, private, liberal arts school. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.