Association for Behavior Analysis International

Five adults diagnosed with autism spectrum disorder (ASD) participated in a 7-week summer training program that targeted conversation skills using computer-based training (CBT) and role-play with peers. Computer-based training provided the definition, rationale, and video demonstrations of one or two individualized targets for each participant (e.g., decreasing interruptions; increasing the duration of utterances). Participants then spoke with other participants during 5-min conversations and received delayed feedback from an experimenter who observed the conversations remotely through the internet. The experimenter used covert audio coaching (CAC) to provide immediate feedback if the initial intervention did not improve performance. The experimenter assessed generalization of the skills to novel adults and in novel settings throughout all phases and maintenance of the skills after the training concluded. Results were inconsistent across participants, suggesting that some conversation skills will require more intensive intervention.

Impairments in social-communication for individuals with autism limits the ability to engage in meaningful and socially reinforcing social interactions leading to social isolation and loneliness. Unfortunately, research on effective social interventions for adolescents and adults with autism is limited. Further, typical interventions aimed at supporting social skill acquisition and maintenance while also fostering independence for adolescents and young adults with autism can be challenging and stigmatizing, particularly given the need for the close proximity of the interventionist. Yet, few studies have capitalized on the use of covert audio coaching (CAC) to deliver evidence-based practices within a natural setting. This study employed a multiple-baseline design across participants to evaluate the impact of CAC with prompting to increase social communication skills in typical social setting for four young adults with autism. Implementation of CAC resulted in increases in the targeted social skill(s) for all participants. Additionally, social validity measures indicate the intervention was viewed favorably by participants. Challenges as well as implications for practice and future research will be discussed.

The purpose of the study was to evaluate effects of a multicomponent training package (scripted self-questions, modeling, and feedback) to establish appropriate stimulus control over social initiations. Participants were three college-age males diagnosed with ASD who were referred to the study based on a history of either initiating or avoiding conversations at socially inappropriate times. First, participants and a sample of typically developing peers completed assessments to inform the development of target scenarios and standards for scoring correct and incorrect performance. Second, we taught participants a self-questioning technique to classify written vignettes as either appropriate or inappropriate conditions to initiate a conversation. Finally, we assessed participants’ initiations under contrived (e.g., scripted interaction between trained research assistants) and naturalistic conditions (e.g., group activity in class). Results indicated that vignette-based training was efficacious for teaching participants to classify untrained written scenarios with preliminary evidence of generalization to interactions with faculty and peers.