Evaluating Predictive Validity by Using Different Scales of the Stanford Achievement Test for the Hearing Impaired

Abstract

It has been suggested that percentile ranks on the subtests of Stanford Achievement Test for the Hearing Impaired yield stronger predictive validity coefficients for the Performance Scale of the Wechsler Intelligence Scale for Children (Revised). This issue of the best criterion scale was re-examined in this study. The predictive validities were evaluated in relation to the two scales of six SAT-HI subtests and three sample sizes. Contrary to previous suggestions, the results indicated that scaled scores of SAT-HI subtests should be used in predictive validity investigations. Scaled scores were found to be consistent and efficient; that is, the validity coefficients approached the true coefficients with increasing sample size.