Psychometric Evaluation of the Swedish Version of the Child Behaviour Check List (CBCL) 1.5-5.

Aim: Examine the psychometric properties of the Swedish Child Behaviour Check List (CBCL) 1.5-5.

Methods: CBCL 1.5-5 responses were collected from 1156 mothers about their 18-month-old children in a population-based study. Representation was analysed by comparing the study sample with two Swedish national samples; scores were reweighted to adjust for selection bias. Confirmatory factor analysis was done using Robust Maximum Likelihood (MLR) estimation on CBCL 1.5-5 syndrome scales and Diagonally Weighted Least Squares (DWLS) estimation for item level. Test-retest reliability was examined with the Pearson (r) coefficient in a subsample (n = 80). Convergent validity was tested between CBCL 1.5-5 and the Early Childhood Behaviour Questionnaire (ECBQ) Very Short Form (n = 826).

Results: Internal consistency of the CBCL scales varied, McDonald's Omega (ω) 0.5-0.9. Confirmatory factor analysis showed good factor fit for seven factors. Test-retest reliability for the Total Problem Score (TPS), Internalising Problems Score (IPS) and Externalising Problems Score (EPS) scales after 4-12 months was r = 0.53, 0.43 and 0.53. Convergence validity was tested between the CBCL 1.5-5 EPS scale and the Early Childhood Behaviour Questionnaire (ECBQ) Effortful control scale (r = -0.51), and the CBCL 1.5-5 IPS scale and ECBQ Negative Affectivity scale (r = 0.45).

Conclusion: The Swedish translation of CBCL 1.5-5 showed good psychometric properties.