Informant Agreement in ASD: Comparisons to Intellectual Disability

Background: When assessing psychological and social functioning, the use of multiple informants is critical to obtain a comprehensive clinical picture. Differing environments, biases and expectations often influence reports from different informants. While agreement among informants has been widely studied in typically developing youth, research on this topic in youth with autism spectrum disorders (ASD) is limited.

Objectives: This study examined informant agreement in youth with ASD using a meta-analytic strategy while making comparisons to informant agreement in youth with intellectual disability (ID).

Methods: Following a comprehensive literature search, a meta-analysis was conducted to determine the average agreement among pairs of informants on rating scales that assess externalizing problems, internalizing problems and/or social skills.

Results: A total of 45 articles were located; 31 reported on informant agreement in youth with ASD and 14 reported on youth with ID. In youth with ASD, agreement on externalizing problems (weighted average r = .41) was comparable to agreement on social skills (weighted average r = .39), but significantly higher than agreement on internalizing problems (weighted average r = .28). Similar results were found for youth with ID (externalizing weighted average r = .44; internalizing weighted average r = .36; social skills weighted average r = .37). Within each behavior type and each informant pair, significant differences did not exist between youth with ASD and youth with ID. In youth with ASD, parents and teachers showed significantly lower agreement on internalizing problems (weighted average r = .21) than did parents and self-report (weighted average r = .48). Pairs of parents showed significantly higher agreement on social skills (weighted average r = .77) in ASD than did parents and teachers (weighted average r = .34), parents and self-report (weighted average r = .45) or teachers and self-report (weighted average r = .39). There were no differences among pairs of raters for externalizing problems in ASD.

Conclusions: Informant agreement in youth with ASD is consistent with that of youth with ID. Informants show higher agreement on externalizing problems and social skills than internalizing problems. Additionally, informant agreement on internalizing and externalizing problems as well as social skills in youth with ASD and ID is comparable to that reported in typically developing youth. Self-reports from youth with ASD or ID show agreement with parent and teacher reports similar to that found in typically developing youth, suggesting this is a potential source of clinical information. These results further highlight the importance of multiple informants in assessing behavior problems and social skills as each informant provides different but important information.