Agreement between mothers’, fathers’, and teachers’ ratings of behavioural and emotional problems in 3–5-year-old children

Abstract

Background

The Strengths and Difficulties Questionnaire (SDQ), a valid and reliable instrument for measuring children’s mental health, is available in parent- and teacher versions, making it an ideal tool for assessing behavioural and emotional problems in young children. However, few studies have evaluated inter-parent agreement on the SDQ, and in most studies on SDQ agreement, parent scores are either provided by only one parent or have been combined into one parent score. Furthermore, studies on SDQ inter-rater agreement usually only reflect degree of correlation, leaving the agreement between measurements unknown. The aim of the present study was therefore to examine both degree of correlation and agreement between parent and teacher SDQ reports, in a community sample of preschool-aged children in Sweden.

Methods

Data were obtained from the Children and Parents in Focus trial. The sample comprised 4,469 children 3–5-years-old. Mothers, fathers and preschool teachers completed the SDQ as part of the routine health check-ups at Child Health Centres. Inter-rater agreement was measured using Pearson correlation coefficient and intraclass correlation (ICC).

Results

Results revealed poor/fair agreement between parent and teacher ratings (ICC 0.25–0.54) and good/excellent agreement between mother and father ratings (ICC 0.66–0.76). The highest level of agreement between parents and teachers was found for the hyperactivity and peer problem subscales, whereas the strongest agreement between parents was found for the hyperactivity and conduct subscales.

Conclusions

Low inter-rater agreement between parent and teacher ratings suggests that information from both teachers and parents is important when using the SDQ as a method to identify mental health problems in preschool children. Although mothers and fathers each provide unique information about their child’s behaviour, good inter-parent agreement indicates that a single parent informant may be sufficient and simplify data collection.