Abstract
Instructional procedures for teaching change computation skills were developed and evaluated on three mentally retarded adolescents. A multiple baseline across subjects combined with the essential features of a multiple baseline across responses was employed. Four response classes, each defined by a different manner of computing change, were taught. Instructional techniques involved: (a) training problems in which the proper responses were modeled, (b) practice problems in which correct responses were reinforced and incorrect responses resulted in corrective feedback, and (c) review problems in which previously trained response classes were rehearsed prior to instruction on a new class. Retraining was provided when subjects' performance fell below the designated criterion. Shaping, chaining, and modeling were the principal antecedent conditions; information feedback and tangible reinforcers were used as consequences. The results showed virtually 100% performance on a posttest and follow-up test. Idiosyncratic patterns of generalization across response classes were noted.
Get full access to this article
View all access options for this article.
