Examining GRE validity with and without restricted sample range

George Finn (2004)

This research examines the validity of the GRE in predicting faculty ratings of student performance during the first year of a psychology doctorate program. Two distinct subject groups are studied: students whose GRE scores were submitted but not reviewed by admissions personnel (1992-1997) and students whose scores were submitted and reviewed by admissions personnel (1998-2001). Results show greater validities for those students whose GRE scores were reviewed as a component of the admission process, with approximately twenty percent of variance explained by GRE scores. Validities for students whose GRE scores were not reviewed by admissions personnel were modest and of limited predictive utility. A review of moderating variables with particular attention to age is recommended to explain observed differences between groups.