Academic journal article Educational Research Quarterly

Predicting Performance of above and below Average Performers in Graduate Business Schools: A Split Sample Regression Analysis

Academic journal article Educational Research Quarterly

Predicting Performance of above and below Average Performers in Graduate Business Schools: A Split Sample Regression Analysis

Article excerpt

Utilizing a sample of MBA students at a medium sized university in the midwestern portion of the U.S., this study employed split-sample regression techniques to determine whether quantitative and verbal GMAT scores, undergraduate GPAs, age, and gender differentially predicted performance for groups of students with above and below median current grade point averages. Results showed that age was a statistically significant positive predictor of performance for individuals with graduate grade point averages above the median, and quantitative GMAT and undergraduate grade point average were marginally statistically significant positive predictors of performance for those students. Gender and quantitative GMAT were statistically significant negative predictors of performance for individuals with current grade point averages below or at the median. These findings indicate that predictor variables frequently used to predict performance in graduate business programs may differentially predict performance for high and low performers.

Graduate schools of business typically establish minimum admission requirements for applicants (Gayle & Jones, 1973; Paolillo, 1982; Youngblood & Martin, 1982). Frequently, scores on the Graduate Management Admissions Test (GMAT) and performance in undergraduate programs are among the key factors utilized by graduate business programs in deciding which applicants are granted admission into programs (Benson, 1983).

Despite the widespread utilization of GPAs and GMAT scores as key factors in admissions decision processes, these measures (even when combined with other potential predictors of success such as work experience, age, and gender) typically explain only a relatively small percentage of overall variation in graduate performance. Benson (1983), Deckro and Woundenberg (1973), Paolillo (1982), and Remus and Wong (1982) could only explain between 15% and 17% of the variance in graduate grade point average with their predictor variables. Similarly, predictors used by Sobol (1984) explained between 9 and 19 percent of the variance in performance, depending on the sample and methodology used.

Carver, Jr. and King (1994) explored predictors of performance for MBA students in a non-traditional program. Variables in their model explained 21% of the variance in graduate performance. Fisher and Resnick ( 1990) reported that, together, GMAT and undergraduate grade point average together explained less than 8% of the variance in first year graduate performance. Graham (1991) found that GMAT was the only significant predictor of performance for students in his study, explaining 17% of the total variance in performance.

Based on these findings, it appears that GMAT scores and undergraduate grade point averages are valid predictors of performance in graduate business schools, but are able to explain only a small percentage of the total variation in graduate school performance. Because the limited predictive ability of variables used by graduate business programs may result in acceptance of students who may be unsuccessful, and the rejection of students who might have been successful, further study of the factors affecting performance of students in graduate business programs is warranted.

Two possible problems associated with the prediction of performance stem from range restriction in the dependent and independent variables, and selection bias based on the fact that only those applicants accepted to graduate business programs are included in analyses (Wilson & Hardgrave, 1995). Typically, applicants to graduate business schools do not encompass the full range of undergraduate grade point averages and GMAT scores. In addition, grades received in graduate business programs generally cluster between 3.00 and 4.00, resulting in a restricted range for the dependent variable. Finally, those individuals with extremely low grade point averages and GMAT scores would be less likely to apply for admission to graduate business programs to begin with. …

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

Oops!

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.