The study compared the effectiveness of two language intervention procedures, modeling alone versus modeling plus evoked production with four language disordered children (ages five and six). There was no marked difference between the procedures as measured on generalization probes. (Author/DB)