The Validity of Einstein Assessment Subtest Scores as Predictors of Early School Achievement

Abstract

The Einstein Assessment of School-Related Skills was developed to help professionals in education, medicine, and health care identify children who need comprehensive evaluation for school learning problems in grades kindergarten through five. This brief (7- to 10-minute) screening instrument was designed to measure reading, arithmetic, auditory memory, language-cognition, and visual motor abilities. The purpose of this study was to evaluate the validity of Einstein subtest scores as predictors of early school achievement. In the fall of 1988, the Einstein was administered to 796 students in kindergarten through second grade from schools in Cleveland, Ohio and the Bronx, New York. Multiple regression analyses were used to identify those subtests that made significant, independent contributions to the prediction of subsequent achievement, with achievement marked by teacher ratings, end-of-year grades, and standardized test scores. Results showed that a single subtest generally predicted as well as the total test, that the best combination of subtests was always an abridgement of the full battery, and that the most powerful predictors tended to be reading measures. Which single subtest or combination predicted best was modulated by both grade level and achievement criterion. Implications of these results for research, clinical practice, and test revision are discussed.