Relative Effectiveness of High School Grades for Predicting College Grades: Sex and Ability Level Effects

Abstract

The effectiveness of standardized test scores (relative to measures of high school performance) for predicting college grades of black students have been disputed.1 As a consequence, some colleges and universities have omitted the use of standardized test scores in their admission procedures in the belief that such scores are not predictively valid for black and other minority students. However, there is evidence that standardized test scores are not only valuable predictors of college grades for black students, but in some instances (particularly for black males), may be more predictively valid than high school rank or average.2 Reasons as to why high school rank (HSR) or average (HSA), traditionally the best single predictors of college grades, do not consistently make greater contributions (relative to standardized test scores) in the prediction of college grades for black students are speculative. Perhaps, sources of unreliability and invalidity of grades are more varied in predominantly black high schools (particularly in overcrowded inner-city schools), especially for males, than they are in the typical predominantly white school. However, there is evidence to suggest that the superiority of standardized tests over HSR or HSA in predicting college grades or attendance also occurs in white male student subgroups with modest