|The development of mathematical models for the prediction of number of undergraduate applications and new student enrollment for colleges and universities: a systems approach using multiple linear regression analysis|
Edward James Yaw
Uploaded by: Pocket Masters
Pockets: Gottesman Libraries Archive, Historical Dissertations
Tags: Admission, Mathematical models, System analysis, Universities and colleges
THE DEVELOPMENT OF MATHEMATICAL MODELS FOR THE PREDICTION OF NUMBER OF UNDERGRADUATE APPLICATIONS AND NEW STUDENT ENROLLMENT FOR COLLEGES AND UNIVERSITIES - A SYSTEMS APPROACH USING MULTIPLE LINEAR REGRESSION ANALYSIS
The primary objective of this research investigation was the development of mathematical models capable of predicting Number of Undergraduate Applications Received and Number of New Students Enrolled for colleges and universities in the United States.
Five hundred twenty-five (58%) of the 900 member institutions of the National Association of College Admissions Counselors returned the research instrument designed by the investigator to elicit the data for the investigation. Using the Systems Approach as a theoretical framework, the data were analyzed through a stepwise selection approach to multiple linear regression.
The final mathematical models were cross-validated, and proved to have high predictive power. Specifically, the final model for the prediction of Number of Undergraduate Applications Received accounted for 88.3% (RSQ = .8834) of the total variance of the criteria. Similarly, the final model for the prediction of Number of New Students Enrolled accounted for 93.3% (RSQ = .9334) of the total variance of this criteria.
The variables which made significant independent contributions to the prediction of Number of Undergraduate Applications received were: Per Cent of Applicants Accepted in the previous year, Prestige Index -- derived by the investigator, Total Admissions Budget per New Student Enrollment Objective, Professional Salary Budget, and Number of Field Representatives.
The variables which made significant independent contributions to the prediction of Number of New Students Enrolled were: Operating Budget of the Institution, Per Cent of Accepted Students Enrolled in the previous year, Freshman/Transfer Mix, Prestige Index, Total Admissions Budget, Total Admissions Budget per New Student Enrollment Objective, Professional Salary Budget per Staff Member, Educational Preparation of the Professional Admissions Staff, and Number of Secondary School Programs held on Campus.
While the models developed proved to be highly predictive of the two primary criteria, the final results suggested that the factors most highly related to the criteria are generally beyond the direct control of those responsible for the Admissions System -- the practicing Admissions Officer. Further, these results suggested that additional recruitment efforts and allocation of additional financial resources to the Admissions System might prove futile at best.