Association for Behavior Analysis International

Traditional approaches to behavior analytic assessment and subsequent treatment of language-limitations have been grounded in Skinnerian Verbal Operant theory. A limitation of Skinners approach was a reliance on exposure to direct contingencies to promote the development of verbal repertoires. Recent advances in Relational Frame Theory provide an alternative approach to language development that does not rely on exposure to direct contingencies for each skill topography, rather a history of reinforcement for relational verbal response classes (e.g., sameness, comparison, etc.). The symposium will compare the convergent validity of assessments based on Skinnerian verbal operant development with those based on Relational Frame Theory and conventional tests of intelligence. As well, we will compare procedures based on a direct-contingency approach to those based in relational training procedures, as described in the PEAK relational training system. Finally, we will showcase psychometric scoring properties associated with the PEAK scoring system and compare this metric to conventional scoring strategies.

Assessing the verbal operant repertoire of individuals with autism is crucial due to the language deficits experienced by this population, and the need for data-driven and individualized treatment. Although several assessments of verbal behavior are available to behavior analysts, few have demonstrated evidence of validity or reliability, and fewer still go beyond elementary forms of verbal operant behavior to complex relational verbal behavior. The PEAK Relational Training System contains two pre-assessments (PEAK-E-PA, PEAK-T-PA) that are used to provide a metric for evaluating participants? abilities to respond relationally to stimuli in their environment cross-modally, and across each of the relational frame families. We will present data correlating the PEAK-E-PA and the PEAK-T-PA with common assessments of IQ (e.g., WISC-V, WPPSI-IV) with 40 children and adolescents with autism, and will provide comparisons in terms of more traditional verbal behavior assessments (e.g., PEAK-Direct Training, PEAK-Generalization, VB-MAPP). The results have implications for the assessment and subsequent treatment of individuals with autism.

Behavior analysts researching language and learning skills in individuals both with and without disabilities have focused primarily on how to establish and increase derived relational responding. Although previous research has indicated that the ability to derive relations may begin before the age of 2 years, little research to date has looked at how the ability to derive relations develops. That is, the ages at which children might be expected to exhibit the ability to derive symmetrical or transitive relations or identify more complex derived relations such as class mergers is unknown. Recently, a curriculum and assessment tool titled the PEAK Relational Training System has been published which includes several modules designed to incorporate both traditional discrete trial training and training for derived relational responding. With the release of the third module, the PEAK Equivalence Module (PEAK-E), behavior analysts now have a standardized assessment tool with which to examine these differing levels of complexity of relational responding. The current study sought to examine scores on the PEAK-E across children from ages 1-22 in both a typically developing population and a population with autism. The results and implications of these findings will be discussed.