Efficiency in Programming for Generalization: Comparison of Two Methods for Teaching Expressive Labeling to Preschoolers with Developmental Delays

Abstract

An adapted alternating treatments design was used to compare the efficiency of two methods of programming for generalization. Four preschool subjects with developmental delays were taught expressive identification of object pictures. During Condition I, single exemplars for each of two objects were trained to acquisition successively until generalization criterion was met. During Condition II, five exemplars for each of two objects were trained concurrently until generalization criterion was attained. Efficiency was measured by comparing the following sets of data for each treatment condition: (a) number of trials/training sessions through acquisition and generalization criteria, (b) rate of errors during acquisition, and (c) total instructional time through generalization criterion. Overall, data indicate that training exemplars concurrently resulted in both faster acquisition and generalization, although the total percentage of errors was greater.