Academic journal article International Journal of Psychology and Psychological Therapy

Using Arbitrary Stimuli to Teach Say-Do Correspondence to Children with Autism

Academic journal article International Journal of Psychology and Psychological Therapy

Using Arbitrary Stimuli to Teach Say-Do Correspondence to Children with Autism

Article excerpt

Since the 1960's, psychologists have explored the correspondence between verbal and non-verbal behaviors (Risley & Hart, 1968; Israel, 1978). One of the first studies on training correspondence (Risley & Hart, 1968), consisted of three separate experiments to develop correspondence training procedures that would lead to generalization. In the baseline condition, children were provided social reinforcement for vocalizations about what toys they played with earlier in the day. During the first phase of treatment, social and edible reinforcement was made contingent upon the participant saying he/she played with the targeted toy. In the second phase, the reinforcement contingency was changed so the child had to play with the targeted toy and make a verbal report.

Results suggested that verbal correspondence only increased when reinforcement was made contingent upon correspondence of doing, and not just the verbalization of doing (saying). When reinforcement was only contingent upon vocalizing the targeted response, the vocalizations or saying would increase, even if the behaviors, or doing, did not increase. In this way, the verbal behavior of an individual can be brought under the control of non-verbal behaviors (i.e., doing may lead to saying with the proper training).

More recent work has focused on the most effective methods for teaching say-do correspondence and for promoting generalization to different environments and activities. Multiple exemplars, graduated prompting strategies, and errorless learning procedures have been used (Luciano, Barnes-Holmes, & Barnes-Holmes, 2002; Luciano, Herruzo, & Barnes-Holmes, 2001). In most cases, multiple exemplars and errorless learning have been shown to be the most effective at teaching say-do correspondence and promoting generalization to novel situations and behaviors (Hernández López, Rodríguez Valverde, & Luciano, 2011; Luciano et alii, 2002).

Some studies have also provided reinforcement to participants for completing a do-say sequence (Lloyd, 2002). Participants were required to perform an action and provide an accurate report of what they did in order to receive reinforcement. Other studies have focused more on using a say-do sequence where participants promise to perform a behavior and receive reinforcement for doing so (Lloyd, 2002; Hernández López et alii, 2011; Luciano et alii, 2001; Luciano et alii, 2002). For example, a child would have to say "I will play with the blocks," and then play with the blocks, not a different toy, in order to gain reinforcement or praise from a teacher.

It is generally agreed that say-do correspondence can be maintained with reinforcement of appropriate verbal behaviors, but that maintaining generalization depends on other factors, such as the function of the behavior (Luciano et alii, 2001). Luciano et alii (2002) used an errorless learning procedure and multiple exemplars to train say-do correspondence in developmentally delayed children. The use of multiple exemplars, along with referent prompts, helped to promote generalization of say-do correspondence. By using a variety of training examples, Luciano et alii (2001) were also able to promote generalization of verbal correspondence while avoiding location or consequence biases. The researchers chose behaviors that had similar functions, which helped to promote generalization of say-do correspondence and suggested that this type of correspondence is a type of rule governed behavior (Luciano et alii, 2001).

Hernández López et alii (2011) studied generalization using behaviors that were not specifically trained, including a small percentage of say-do relationships that were the opposite of the ones taught in the training phase. By including simple stimulus classes as discriminative stimuli during correspondence training children were able to show both correspondence and non-correspondence for a wide variety of behaviors that were never directly trained (Hernández López et alii, 2011). …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.