Testing the Validity of GRE Scores on Predicting Graduate Performance for National and International Engineering Students

Abstract

Numerous research studies have examined the validity of GRE scores in predicting graduate success, however, some limitations still exist. This study targeted graduate engineering programs and investigated the validity of GRE scores in predicting graduate engineering GPA (GGPA). In addition, the differences in the validity of GRE scores between American and international students and between masters’ and doctoral students were compared. The GRE’s incremental predictive ability over undergraduate GPA (UGPA) and TOEFL scores was examined. Data were obtained from 1083 students from the engineering programs in a large, comprehensive midwestern university. Results indicated that GRE was useful in forecasting GGPA of graduate engineering students. The GRE scores explained more criterion variance for American students than for international students, but statistically significant differences were only found when GRE-Quantitative predicted GGPA. The GRE-Verbal and GRE-Quantitative scores had different patterns in predicting graduate grades for master’s and doctoral students. UGPA was found to be a very strong predictor, and TOEFL scores were significantly correlated with the criterion variables. GRE scores, however, were found to have significant incremental validity over UGPA and TOEFL scores. TOEFL scores were less able to make predictions.