Using a Blocked-Trial Procedure to Teach Identity Conditional Discrimination in Individuals with Autism: Effect of using Three Sample Stimuli

Abstract

The blocked-trial procedure is an effective procedure to teach matching-to-sample skills. This procedure usually uses two comparisons and two sample stimuli that are alternated in blocks of consecutive trials. Such blocks are reduced with regard to the number of trials until samples are presented in random order. However, the procedure can be considered inefficient because it requires a large number of trials to teach two relations only. The present study investigated the effects of increasing the number of samples to three in an identity matching task. Five individuals with autism participated in the study. The sessions always comprised 24 trials. As an initial step, the same sample was displayed in all of the trials in a session. If the participants showed accurate performance, then the three sample stimuli were introduced in progressively smaller blocks of consecutive trials: 24 to 8 trials, 8 to 4 trials, 4 to 2 trials, and finally in random order. When the participants learned new identity relations, they were given an generalized identity matching test. All of the individuals learned the matching task for at least two stimulus sets, although alternative procedures were required to teach some of the participants. Four participants showed generalized responding. The blocked-trial procedure with three samples effectively taught three conditional relations simultaneously, which is important for effective teaching and avoiding incoherent stimulus control topographies. The discussion addresses issues about the procedures, programming conditional discrimination learning, and generalized identity matching in individuals with developmental disabilities.   Keywords: blocked-trial procedure; conditional discrimination; generalized identity matching; autism.