The Accuracy of Temperament Ratings in Late Childhood and Early Adolescence and Their Relation to Behavior Problems: An Analysis of Parent, Child, and Teacher Agreement

Interrater agreement on child temperament and personality is often found to be low to moderate. While some interpret this finding as a psychometric problem, others argue that it may instead reflect the distinct contexts and perspectives that shape each informant’s perception of the child. To gain a better understanding of this issue, we adapted the parent version of the Integrative Late Childhood Temperament Inventory (ILCTI) for use in children and teachers and examined the agreement between parents, children, and teachers. We collected 280 self- and teacher-ratings of the ILCTI from 9-14-year-old Austrian children and young adolescents. We also obtained parent ratings for 110 children and a second teacher rating for 57 children. Results on psychometric properties supported internal consistency reliability and criterion validity (meaningful associations with internalizing and externalizing problems and school performance) for all three versions. Interrater agreement across all items of the ILCTI (as assessed by dyad-centered agreement), and on the six individual temperament dimensions was highest among teachers, followed by parent-child, parent-teacher, and teacher-child agreement. Interestingly, low parent-teacher agreement was associated with more behavioral and emotional problems in children. Possible explanations for differences in the extent of interrater agreement among rater groups are discussed.