Do We Need Multiple Informants When Assessing Autistic Traits? The Degree of Report Bias on Offspring, Self, and Spouse Ratings.

This study focused on the degree of report bias in assessing autistic traits. Both parents of 124 preschoolers completed the Social Communication Questionnaire and the Autism-spectrum Quotient. Acceptable agreement existed between mother and father reports of children's mean scores of autistic traits, but interrater reliability for rank-order correlations was only fair. No evidence was found for report bias regarding parent-offspring autistic traits. However, adult autistic ratings were strongly biased: spouse-ratings were higher than self-ratings, correlations were only fair when both parents reported about the same person, and resemblance was higher for reports from the same person than for spouses' separate self-reports. It is advisable to involve multiple informants when assessing autistic traits, and to use procedural and/or statistical remedies to control for report bias.