Classroom observation code for hyperactive children: A replication of validity.

This study was designed to replicate the validity of a modified State University of New York at Stony Brook classroom observation code to distinguish normal children from children evaluated as hyperactive by their teachers. All previously obtained code results were cross-validated: The code category cutoff scores clearly distinguished hyperactive from normal children; the combination of categories into dyads yielded better discrimination between hyperactive and normals than did single categories; the behavioral dyad of interference and off task was the most discriminating. Other findings pertinent to the evaluation of hyperactive children were replicated. Namely, age was found to be inversely related to level of observed classroom behavior, and hyperactive children had greater day-to-day within-subject variability than did normals. In addition, no observer effect on the children's behavior was found. This replication gives strong support to the reliability and discriminative validity of the code and documents its utility in studies of hyperactive children.