A Comparison of Teaching Procedures to Teach Auditory-Visual Matching-to-Sample Tasks

We evaluated 3 procedures for teaching auditory-visual conditional matching-to-sample tasks using a parallel treatments design. An adolescent boy with autism served as the participant, and 2 comparisons were conducted. Comparison 1 evaluated effects of three teaching procedures: error correction, gestural prompt, and picture prompt. A different set of 3 auditory-visual relations was assigned to each procedure. Baseline accuracy ranged from 22 to 44%. The mean accuracy scores for the last 3 sessions of the training phase were 100% for the picture prompt condition, 85% for the error correction condition, and 7% for the gestural prompt condition. Comparison 2 evaluated effects of the gestural prompt and picture prompt procedures. Baseline accuracy was 33% for both sets of relations, and the mean scores for the last 3 sessions of training were 93% for gestural prompt and 0% for picture prompt. The inconsistent findings between the two comparisons are discussed.