Correspondence of Parent and Trained Observer Reports of Social Skills and Autism Symptoms

Background:

It has been documented that in the child population, discrepancies often exist among different informant’s ratings of child psychopathology and these discrepancies have an impact on assessment, classification, and treatment. Very little research has examined rater discrepancies, particularly in populations of children with developmental disorders.

Objectives:

The first objective of the study was to determine the consistency of parental reports of autism symptoms and social skills. This study also aimed to investigate the level of disparity between parent and trained observer reports of social skills and autism symptoms.

Methods:

The sample consisted of sixty-eight elementary school-aged children who were participating in a randomized treatment trial (M = 9.9 years, SD = 1.8) diagnosed with high functioning autism spectrum disorder (IQ above 85).

The parent report measures examined were the Autism Diagnostic Interview Revised (ADI-R) and the Social Responsiveness Scale (SRS). The ADI-R is an audio taped semi-structured interview organized around DSM-IV and ICD-10 diagnostic criteria for autism, and the SRS is a 65-item rating scale parent report measure of children’s social impairments in naturalistic social settings.

The trained observer measures examined were the Autism Diagnostic Observation Schedule (ADOS) and the Playground Observation of Peer Engagement (POPE). The ADOS is a semi-structured autism diagnostic assessment tool and the POPE is a behaviorally based timed interval behavior coding system that measures peer engagement in natural environments.

Results:

Pearson correlations were performed between parent and trained observer ratings. There was a nonsignificant correlation (r = .08, p = .491) between measures of autism symptoms reported by parents (ADI-R) and those rated by trained observers (ADOS). Similarly, a nonsignificant correlation (r = -.175, p = .144) was found between measures of social skills reported by parents (SRS) and those rated by trained observers (POPE). In a secondary analysis, we found a significant correlation (r = .529, p < .001) between the parent report of autism symptoms (ADI-R) and parent report of social skills (SRS). At the subscale level, ADIR-restricted, repetitive, and stereotyped patterns of behavior significantly correlated with SRS-total (r=.264, p=.008), SRS-communication (r=.210, p=.035), SRS-motivation (r=.239, p=.016) and SRS-mannerisms (r=.382, p=.000).

Conclusions:

Parent and trained observer reports of autism symptomatology and social skills were not found to be consistent with one another. Greater child impairment reported by parents did not equate to worsened performance ratings granted by clinicians. Our findings provide support for the idea that parent report is not in line with eventual diagnostic outcome. Our results of the comparison between SRS and ADI-R indicate that parent report of ASD symptomatology and behaviors remained consistent across report type. Specifically, the information gathered during the semi-structured interview was reflected in the parent self-report measure. It is important to further investigate the lack of a relationship between parent and trained observer reports, given the need for accurate information to create clinical profiles and deliver effective treatments.