Catching sight of children with internalizing symptoms in upper elementary classrooms

Abstract

Teachers play a crucial role in the assessment of children's internalizing symptoms but may not always succeed in accurately identifying such symptoms in class. Using a multilevel structural equation modeling (MSEM) approach, this study aimed to explore teacher and child characteristics that may explain measurement bias in teachers' ratings of internalizing symptoms at the between- and within-teacher level. Upper elementary school teachers (N = 92, 74.9% female) filled out the Strengths and Difficulties Questionnaire, Student-Teacher Relationship Scale, and Student-Specific Teacher Self-Efficacy Scale for randomly selected children (N = 690, 50.5% girls, Grades 3-6) from their classrooms. Participating teachers and children also responded to several background questions. Multilevel SEMs suggested that teachers' self-efficacy beliefs toward, relationship experiences with, and externalizing symptom ratings of individual children affected their ratings of these children's internalizing symptoms at the within-teacher level. Specifically, given equal levels of internalizing behavior, teachers were likely to systematically under-identify symptoms of anxiety and over-identify bullying for children with more externalizing behavior and conflictual relationships, or in circumstances where teachers had lower self-efficacy. Children with high levels of closeness received systematically higher ratings on somatic complaints and lower ratings on solitary behavior and peer problems. At the between-teacher level, less experienced teachers were more likely to over-identify symptoms of worries than were more experienced teachers, given equal levels of internalizing symptoms. As such, these findings extend the limited body of evidence on children's internalizing symptoms in upper elementary school.

Keywords: Background characteristics; Internalizing behavior; Measurement bias; Student–teacher relationships; Teacher self-efficacy.