Date of Award


Degree Name

Doctor of Philosophy



First Advisor

Dr. C. Richard Spates

Second Advisor

Dr. Linda LeBlanc

Third Advisor

Dr. Galen Alessi

Fourth Advisor

Dr. Stephanie Peterson


Recent estimates of the incidence and prevalence of autism spectrum disorders (ASDs) indicate substantial increases over the past 20 years (Centers for Disease Control and Prevention, 2009; Fombonne, 2009; Matson & Kozlowski, 2011; Schreibman & Koegel, 2005; Wing & Porter, 2002), in part because of the availability of significantly improved diagnostic assessment instruments (Lord & Corsello, 2005; Schreibman, 2005). In highly structured research settings, some of these diagnostic instruments correlate well with each other; however, few studies have examined the relation between these diagnostic tools in traditional clinical practice (Mazefsky & Oswald, 2006; South et al, 2002). This study examined archival client data for 77 cases from two outpatient clinics serving children with suspected ASD to investigate the psychometric properties of diagnostic assessments when used in clinical practice. The obtained psychometric properties were compared to published psychometric properties from research sites. Obtained reliability and validity measures were much lower for the current study than those in published psychometric studies. This study also examined whether several published findings in the ASD literature were replicated in the community sample. Unlike some previously published studies, no significant differences existed between the receptive and expressive language scores for those children with a final diagnosis of ASD and those with either a non-ASD diagnosis or no diagnosis. The current study did not find significant differences between the adaptive behavior composite for children with a final diagnosis of ASD and those with either a non-ASD diagnosis or no diagnosis. Similar to published studies (Mazefsky & Oswald, 2006; South et al., 2002), this study found that the mean GARS score and the mean GADS score for those with a final diagnosis of autism/PDDNOS or Asperger's, respectively, underestimated the probability of being on the autism spectrum when compared to more extensive assessment using direct observation tools. These results highlight the importance of (a) utilizing the multimethod- multi-trait assessment process in community, clinic-based settings to safeguard against overgeneralization from test scores alone, and (b) evaluating psychometric properties of diagnostic instruments in their most commonly used settings in addition to extant research protocols.

Access Setting

Dissertation-Open Access