Utility of the Social Responsiveness Scale- Parent Report (SRS-Parent) as a Diagnostic Tool for Autism Identification

Thesis (Ph.D.)--University of Washington, 2016-06Rating scales are often used as part of the evaluation process to diagnose autism spectrum disorder (ASD). Rating scales that are modeled after the experiences and understanding of the Caucasian American race may not reflect the unique experiences of individuals from other races or ethnicities. If parent ratings do not uniformly identify the ASD symptoms of children from varying racial and ethnic backgrounds, it likely will impact the diagnostic process, which may contribute to the known disparity in identification rates by race and ethnicity. The Social Responsiveness Scale (SRS) is one rating scale that can be used by parents and teachers. Although there have been many studies addressing the reliability and validity of the SRS, there is a dearth of available research that conducted similar analyses by racial/ethnic group. This study endeavored to identify the appropriateness of using the Social Responsiveness Scale- Parent Form (SRS-P) by evaluating its reliability, criterion validity, and structural validity across racial and ethnic groups using children identified with ASD in the Simons Simplex Collection (SSC) dataset. When analyzed independently, the Autistic Mannerisms and Social Communication scales were the only reliable scales found when using the total sample and when using the Caucasian American, African American, and Asian American groups. For the Mixed Race and Hispanic groups, Social Communication was the only reliable scale. A linear regression found that only the total SRS-P score and the Autistic Mannerisms scale, which targets stereotypical behaviors and highly restricted interests, significantly predicted the ADOS-CSS. As a whole, the SRS-P does not appear to differ in its ability to identify severity of symptoms of ASD by racial/ethnic group. An exploratory factor analysis done by total SSC sample and by racial/ethnic groups within the sample suggests that the SRS-P is most appropriately considered as a single-factor instrument. Rather than use the treatment subscales, intervention goals may be best created based on the endorsement of individual items. School psychologists need racially and ethnically fair instruments to use with their school populations. As the SRS-P performed similarly regardless of the race or ethnicity of the child/parent and the total score significantly predicted the ADOS-CSS, this particular instrument may be a valuable resource to use with children from diverse racial and ethnic backgrounds. Recommendations for use of the SRS-P and implications for future research are included