Predictive Validity of the Graduate Record Examination Advanced Psychology Test for Grade Performance in Graduate Psychology Courses
House, J. Daniel, Johnson, James J., College Student Journal
Graduate Record Examination (GRE) scores are often considered during the admissions process for prospective graduate students and there have been several assessments of the predictive validity of the GRE for graduate students in psychology. The purpose of this study was to investigate the predictive validity of the GRE Advanced Psychology Test for students' grade performance in selected graduate courses. Academic records were evaluated for 236 graduate students in professional psychology programs. Higher GRE scores were significantly correlated with higher grades in several specific courses. These findings indicate that, in some instances, GRE Advanced Psychology Test scores are significantly correlated with subsequent performance in selected graduate psychology courses.
Graduate Record Examination (GRE) scores are often considered during the admissions process for prospective graduate students (Oltman & Hartnett, 1984). There have been several assessments of the predictive validity of the GRE for graduate students in psychology. Most of these studies have examined the validity of the Verbal (GRE-V) and Quantitative (GRE-Q) sections of the GRE for the prediction of subsequent achievement outcomes. For instance, GRE scores have been found to be significant predictors of cumulative grade performance (House, Gupta, & Xiao, 1997), grades in specific graduate psychology courses (House & Johnson, 1998; Huitema & Stein, 1993), and graduate degree completion (House & Johnson, 1993a). Further, GRE scores were significantly correlated with a multifaceted rating that reflected student progress in a clinical psychology graduate program (Dollinger, 1989). Recent results also suggest that GRE scores were significant predictors of whether American Indian/Alaska Native students completed their graduate degrees (House, 1997). A more limited number of studies, however, have examined the validity of the GRE Advanced Psychology Test and those studies have produced conflicting results. GRE Advanced Psychology Test scores were found to significantly predict the comprehensive examinations performance of students in clinical psychology (Kirnan & Geisinger, 1981) and graduate grade performance (Federici & Schuerger, 1974). House and Johnson (1993b) found that GRE Advanced Psychology Test scores were significantly correlated with the cumulative graduate GPA of students in professional psychology programs.
There are multiple criteria that may be used to assess graduate student achievement, including grades, degree completion, time to degree completion, and examinations performance (Hartnett & Willingham, 1980). While GRE Advanced Psychology Test scores were significant predictors of cumulative graduate GPA, it has been noted that further study is needed to determine if scores would be significantly related to other types of criterion measures, such as grades earned in specific courses (House & Johnson, 1993b). Consequently, the purpose of this study was to investigate the predictive validity of the GRE Advanced Psychology Test for students' grade performance in their graduate courses.
Academic records were evaluated for 236 graduate students enrolled in professional psychology programs at a large midwestern university. The sample was comprised of 101 students in clinical psychology, 76 students in counseling psychology, and 59 students in school psychology. Data were collected for three predictor variables, GRE Advanced Psychology Test scores and the Experimental and Social subscores from the Advanced Psychology Test. In addition, students' grades in ten courses frequently taken by graduate students in these programs were compiled. Course grades were assigned on a 4-point scale (A = 4 and F = 0). Validity coefficients were computed for the relationship between GRE scores and course grades. Given the number of correlations being tested, the Bonferroni approach to multiple significance tests was utilized and a conservative significance level of . …