Principal Component Analysis of the PEAK Generalization Module

Abstract

The current study sought to assess the content validity and internal consistency of the PEAK Generalization Module. Eighty-four children with autism were evaluated using the PEAK Generalization Assessment to ascertain the presence or absence of 184 language and learning skills within the child’s repertoire. Following the assessment, a principal component analysis was run yielding a four-component model of the PEAK Generalization Module. Specifically, components possessing eigenvalues greater than 3 and that had at least one item which was most strongly correlated to them were retained. Items were then sorted into the various components based on their correlation scores within the rotated component matrix generated by the principal component analysis. Following the establishment of the four-component model, the internal consistency of the model was tested using Cronbach’s Alphas which indicated strong internal consistency in the overall PEAK-G Assessment as well as each of the four underlying components. The four components identified include the constructs of Foundational Learning and Basic Social Skills, Basic Verbal Comprehension, Memory, and Advanced Social Skills, Advanced Verbal Comprehension, Reading and Writing, and Basic Problem Solving Skills, and Verbal Reasoning, Problem Solving, Logic, and Mathematical Skills. With the increased demand for evidence-based intervention as the number of individuals with diagnoses of autism rises, the present data provide support for a conceptually systematic behavior analytic approach to the treatment of children with autism.