Examining Agreement in Parent and Teacher Report for School-Aged Children at-Risk for Autism Spectrum Disorder

Background:  Parents and teachers are frequently asked to provide information about a child’s emotional, behavioral, and social functioning when there is a question of possible Autism Spectrum Disorder (ASD). Two commonly used measures include the Achenbach Child Behavior Checklist (CBCL) and the Social-Responsiveness Scale, Second Edition (SRS-2). Parent and teacher report are important factors when considering a possible ASD diagnosis for school-age children. Therefore, it is important to understand the relationship between reports from these informants.

Objectives:  To examine the relationship and inter-rater consistency (ICC) between parent and teacher report for a sample of children at-risk for ASD.

Methods:  Data for this study is from the South Carolina Children’s Educational Surveillance Study (SUCCESS), which is designed to assess the prevalence of ASD through direct screening and evaluation. Children born in 2004 and aged 8-11 at the time of participation were identified as being at risk for ASD via population-based screening using the Social Communication Questionnaire (SCQ). Children at-risk for ASD received full developmental assessments to determine diagnostic status (n=292). For the majority of children (n=247), both parents and teachers completed questionnaires including the SRS-2 and the Achenbach Child Behavior Checklist (CBCL). Data analyses included parent versus teacher predicted means via mixed model approach and estimated ICC.

Results:  In the total sample of children at-risk for ASD, both the Social Communication Index (SCI) and Total Score of the SRS-2 were rated higher by parents than by teachers (LSM difference 6.27 (SE=0.87); p < .0001; 6.09 (SE=0.87); p < .0001, respectively). ICC was fair for the SCI (0.39) and moderate for the Total Score (0.42). Significant differences in parent and teacher ratings on the SRS-2 were observed in children both with and without the diagnostic classification of ASD. In the total sample, parents also rated children higher in internalizing problems on the CBCL (LSM difference 4.42 (SE=0.88); p < .0001, and ICC was fair (0.23). However, no significant differences were found between parent and teacher report of externalizing problems on the CBCL. Among children who screened at-risk but were not diagnosed with ASD (n = 202), parents also rated children higher in internalizing (LSM difference 5.16 (SE=0.96); p <.0001) and externalizing problems (LSM difference 1.85 (SE=0.83); p=0.027) on the CBCL. However, among children diagnosed with ASD (n = 45), there were no significant differences between parent and teacher report on internalizing or externalizing behaviors on the CBCL. ICCs were fair for both internalizing (0.20) and externalizing problems (0.37).

Conclusions:  Results suggest that, on average, parents are more likely to endorse social communication concerns than teachers. Parents also endorse more internalizing and externalizing behavior problems among children who are at-risk but do not go on to be diagnosed with ASD, while there is greater agreement between parents and teachers on these scales for children subsequently diagnosed with ASD.