Association for Behavior Analysis International

Much of the research in training verbal skills to children with autism or language delays is largely centered around establishing a basic repertoire of echoics, mands, tacts, and intraverbals leaving much to be desired. Skinner's own approach to verbal behavior extended far beyond these basic concepts, and more advanced verbal operants included in his analysis have been utilized in language training only minimally. Additionally, behavior-analytic investigations of verbal behavior have since extended into stimulus equivalence and relational frames which both offer many benefits to establishing a verbal repertoire and may have further implications for teaching verbal skills to children with language delays. The PEAK Relational Training System, a verbal behavior and academic curriculum for children with autism that incorporates all of these basic and more advanced learning areas, holds promise in extending the current methods of behavior analytic language training. The present research examines the psychometric validity and reliability of PEAK, as well as barriers, agency satisfaction, and practical utility of the system. Approaches such as these may further allow behavior analysts to promote verbal skills in children with autism that are more generative and more broadly applied to enhance functioning in the natural environment.

Verbal skills curriculums and assessments for children with autism hold value for a practitioner or agency only if the implementation of such a system is feasible, sustainable, and produces the outcomes desired from such a package. This presentation focuses on one community provider's attempts at implementing the PEAK Relational Training System on persons with autism. We will highlight the initial identified barriers to implementation of the system, discuss further training needs for successful implementation, showcase outcomes of children with autism who were engaged with the system, and discuss social validity measures of the PEAK Relational Training System, including reported ratings of implementer and agency satisfaction, feasibility of use, and long term sustainability. Finally, we will present a model of how to attain optimal PEAK outcomes based on the results of mass implementation throughout one service agency with children with autism. Overall, staff reliably reported that the PEAK Relational Training System and assessment were easy to implement, that children were demonstrating visible progress, and that they were highly satisfied with the curriculum and assessment.

Despite the number of assessment protocols available to practitioners in a data-driven field, measures of validity and reliability of these assessments remain a rarity. The purpose of the present research was to examine psychometric properties of the PEAK Relational Training System, a comprehensive assessment/curriculum for verbal and academic skills. First, 13 children with autism completed the PEAK assessment and the Peabody Picture Vocabulary Test, a measure frequently utilized to assess verbal skills in children, before and after training using the PEAK curriculum. A Pearson correlation showed a significant (p < .01), strong relationship between the measures at pre-training (r=.9) and post-training (r=.89), establishing convergent validity for the PEAK assessment. Next, 36 students with autism completed both the PEAK assessment and the Receptive and Expressive One-Word Picture Vocabulary Tests. Strong correlations were observed between scores on PEAK and the receptive test (r=.834), and the expressive test (r=.824), further supporting the convergent validity of the PEAK assessment. To investigate social validity, ABA implementers rated the PEAK assessment and VB-MAPP on ease and length of implementation, and the likelihood of future use. Results indicated that practitioners found PEAK easier and quicker than the VB-MAPP, and 66% reported a preference for PEAK.

The current presentation will showcase how a normative sample of typical developing school children were captured using the PEAK Relational Training System Assessment. Over 100 students of various age ranges with no known or documented disabilities were administered the Direct Training Assessment of the PEAK Relational Training System and their aggregate scores were compared to children with autism across similar age ranges. Each child was administered the PEAK Direct Training Assessment which evaluated the presence or absence of 184 skills which focused on academic, verbal, cognitive, and social development shown to be aligned with Common Core Standards, Peabody Picture Vocabulary Test, and IQ. Performances were then evaluated and compared to children with Autism exposed to the same assessment. This normative technique provides for a comparison of children with autism to typically developing peers, and also expands beyond developmental milestones hypothesized to be constant across children within a given age group.