An examination of the predictive validity of the keymath diagnostic arithmetic test and the wide range achievement test in exceptional children

Thirty-seven Title I elementary school children were administered the Arithmetic subtest of the Wide Range Achievement Test and the KeyMath Diagnostic Arithmetic Test. One year later, the Metropolitan Achievement Test was administered. While correlations between the Wide Range and the Metropolitan did not reach significance (r =.17), the KeyMath demonstrated a substantial relationship (r -.63). Reasons for such differences in correlations are discussed. It is suggested that the reliability and predictive validity of the KeyMath be extended to different ages and populations.