Association for Behavior Analysis International

Seven children diagnosed with autism participated in this study. All participants can respond to simple WH questions, for example, stating their names or identifying objects and actions when being asked. However, they can only engage in multiple verbal exchanges with verbal prompts. In this study, trainings were provided for all participants to engage in multiple verbal exchanges with specific topics. Participants were also taught to talk about specific topics according to the interests of their conversational partners. There are total of five training phases. First, participants were taught to tact immediate family members, teachers, and friends. Then, they were taught to match topics of interests to each person. Then, they were taught to generate five statements about each topic. Finally, role-play and generalization of intraverbal responses were introduced. Result shows that all participants were able to reach mastery criteria of 80% independence.

Most children with autism lack the natural ability to develop speech. They have to be trained using specific teaching models to develop speech. There is substantial research for increasing vocalizations in children by using stimulus stimulus pairing (Normand & Knoll; 2006), automatic reinforcement procedures (Sundberg et al; 1996), and manual sign mand training with prompt delay (Attanasio et al; 2012 ). However there is not enough research using the same methods for inducing speech in non vocal children. This study included 3 non vocal children with a diagnoses of autism aged 4 years, 7 years and 13 years. A multiple baseline design across participants verified the effectiveness of the procedure which included teaching the child to mand using signs with stimulus stimulus pairing. Once communication through sign was achieved to independence under motivating operation, the protocol was modified to a 2 sec time delay in vocal prompting. Results in all three children exhibited acquisition of sign mands followed by emergence of vocal mands across environments.