Prediction of Special Education Need from Preschool Screening Data

Predictive validity of two measures used for early detection of learning problems was investigated using educational group placement as a criterion. Test scores for 240 children were collected prior to kindergarten entrance. Educational placement was examined at the end of grade one based upon assignment to one of four groups (regular education, special education, rescreen-regular education, and retention). Predictive validty was examined using the Discriminant Analysis method. One linear combination of preschool test scores was found which accounted for 95% of the available variance due to group membership. Comparison of group classification using the function and actual group membership indicated that 65% of the children screened were accurately classified on the basis of screening results. The function was successful in the differentiating children in need of some form of special education from those who did not.