Date of Award
University or Center
Atlanta University (AU)
School of Education
The purpose of this study was to determine the efficiency of the Graduate Record Examination Aptitude Tests (GRE) and the Miller Analogies Test (MAT) as predictors of academic performance for doctoral students in the Department of Psychological Services.
Significance of the Study
This study was expected to increase the body of knowledge in the area of prediction of academic performance and criteria selection for use in doctoral programs.
Four hypotheses were tested. Null hypotheses one. two. and three involved bivariate analyses of the Graduate Record Examination (GRE). Miller Analogies Test (MAT). and academic performance (GGPA). The fourth null hypothesis employed the same variables in a multiple regression analysis.
Subjects The subjects were nineteen doctoral graduates who completed graduate study during the years 1969-1983. in the Department of Psychological Services. School of Education at Atlanta University.
Instruments The Miller Analogies Test (MAT) and the Graduate Record Examination Verbal and Quantitative Aptitude Test scores were utilized in this study.
1. Only the Graduate Record Examination Verbal (GRE-V) was significantly correlated with graduate grade pOint average (GGPA). It made the greatest contribution to the prediction of grade point average (GPA) and was the most consistent predictor variable when used with the GRE-Q and GRE-T.
2. The Miller Analogies Test (MAT) was not a predictor of successful academic performance (GGPA).
3. The GRE-V and GRE-T were statistically significantly correlated with the Miller Analogies Test (MAT), however, the Graduate Record Examination Quantitative (GRE-Q) was not statistically significantly correlated with the Miller Analogies Test (MAT).
4. When bivariate correlation coefficients with multiple correlation, using the verbal, quantitative and total Graduate Record Examination (GRE) scores were compared, the bivariate correlation involving the GRE-V provided the only statistically significant correlation. Although this correlation was statistically significant, its predictive power was limited.
Tift, Helen E., "The prediction of academic performance in a doctoral counseling program" (1984). ETD Collection for AUC Robert W. Woodruff Library. 1771.