Factor structure and reliability of the parent-informant strengths and difficulties questionnaire in a Scottish preschool sample

Research Findings: Despite it being a widely accepted measure of social, emotional, and behavioral problems among children and young people, the factor structure underlying the Strengths and Difficulties Questionnaire (SDQ) is contested. Recent research has suggested a need to consider methodological bias in the SDQ, associated with the positive wording of some of the items, as a separate latent construct. Using data from the 2006–2007 Growing Up in Scotland study, we performed confirmatory factor analyses with 1,789 parents’ SDQ reports of their 4-year-old children, estimating and comparing 3 competing factor models: a 5-factor model, 3-factor model, and a 6-factor model incorporating a method factor. Results indicated that the 6-factor solution was the best fitting model. Several items from the original Pro-Social Behaviour and Peer Relationship Problems subscales loaded higher on the method factor than the original, suggesting bias influences across the SDQ domains. Reliability analysis demonstrated poor internal consistency for the individual difficulties subscales but good internal consistency for a Total Difficulties score and moderate internal consistency for the strengths subscale. Practice or Policy: As the SDQ is widely used internationally, further research is required to facilitate interpretation across populations and purposes (i.e., screening, diagnosis, or outcome assessment).