Predicting Placement Success for Applicants to a Full-Time MBA Program

Charles R Johnson, Purdue University

Abstract

Historically, the success of a full-time MBA program has been judged directly or indirectly on the program’s ability to demonstrate strong placement outcomes for its graduates. Placement outcomes are used by prospective students when determining where to attend school and are often a significant component of various MBA program rankings. This dissertation explores possible relationships between student variables known at the time of admission to a full-time MBA program and subsequent placement outcomes for those same students. Admissions and placement data comprising twelve cohorts of the full-time MBA program at Purdue University’s Krannert School of Management were examined to determine if a model predictive of placement outcomes, job acceptance and starting base salary, could be developed from applicant data including undergraduate grade point average, GMAT scores, and months of pre-MBA work experience. Potential placement. Impacts of gender and U.S. residency were also examined. The study also explored the potential relationships between academic performance in the MBA program and placement results. Regression analysis was used to explore relationships to starting base salary as reported by MBA graduates over the period. Similarly, logistic regression was used to explore relationships between admission variables and a graduate’s reported acceptance of employment within three months of graduation. The research findings did not produce a model predictive of placement success, suggesting that factors not captured or not coded at the time of admission may have a more significant impact on placement success than the traditional factors used to make MBA admission decisions. The study did find statistically significant correlations between independent admission variables and placement outcomes, however. Pre-MBA work experience was generally positively correlated with both starting salary and job acceptance. Not having U.S. citizenship or residency was shown to be negatively correlated to both starting salary and job acceptance. Examining the potential relationship between academic performance in the MBA program produced an interesting dichotomy. Year one grade point average was statistically significantly positively correlated to base salary, while year two grade point average was statistically significantly negatively correlated to salary. A similar dichotomy was not found for job acceptance outcomes. The findings suggest that traditional factors evaluated at the time of MBA admission may be inadequate for identifying future placement success and that the development and capture of new admissions criteria may be needed. Though the findings from this study are impacted by the unique characteristics of a particular MBA program and cannot be generalized, the framework of the study could be used by other MBA programs attempting to predict placement outcomes. Implications for practitioners and for future research are discussed.

Degree

Ph.D.

Advisors

Knupfer, Purdue University.

Subject Area

Higher Education Administration

COinS