Date of Submission
Doctor of Psychology (PsyD)
Robert A. DiTomasso, Ph.D., ABPP
George McCloskey, Ph.D.
Daniel H. Ingram, Psy.D.
Kathryn Gipe, Psy.D.
Early identification of autism has received national attention and can lead to positive treatment outcomes. Research has demonstrated that qualitative impairments in social and communicative behaviors can be detected within the first two years of life. Unfortunately, many children with autism will not receive an autism diagnosis until they enter preschool. Professionals within school settings have not often received extensive training on assessment of autism spectrum disorders. Expectations, however, are that school-based teams can identify students who demonstrate characteristics associated with autism and refer those students for a comprehensive autism evaluation. Currently, autism screening tools are limited to rating scales completed by parents or teachers. Although rating scales have value in the screening process, they should not be relied upon solely to determine whether or not further assessment is warranted. Instead, rating scales should be used in conjunction with direct observation in determining if an intensive autism evaluation is necessary. Because direct observation can be subjective and reliant on the skills and experience of the observer, a tool is needed to guide and quantify observations. The current research study provided a retrospective analysis of such an observational screening tool used to observe students enrolled in a preschool program within the Christina School District. The first research question examined the relationship between scores on the observational checklist for students with typical development, developmental delays, and autism. Students with autism scored significantly lower than students with typical development and also lower than those with developmental delays. The second research question examined the effectiveness of individual checklist items at differentiating between the identified groups. Although all items effectively discriminated autism from typical and developmentally delayed peers, removing the two least effective items increased the overall sensitivity, positive predictive validity and negative predictive validity of the checklist. The third research question investigated the relationship between developmental level and overall checklist score. Developmental level, characterized as “high” (cognitive score of 85 or greater) or “low” (cognitive score of 84 or lower) did not correlate with performance on the checklist. Regardless of developmental level, the checklist correctly identified children within the autism group, based on a cut-score lower than 10 (on the 12-item checklist).
Heim Elberson, Mary, "Improving Validity of Referrals for Comprehensive Autism Evaluation : A Tier-2 Direct Autism Observation Checklist" (2010). PCOM Psychology Dissertations. 153.