Discriminant validity of the child behaviour checklist CBCL-4/18 in German samples

Eur Child Adolesc Psychiatry. 2001 Dec;10(4):240-7. doi: 10.1007/s007870170013.

Abstract

This study examined the discriminant validity of the German version of CBCL in two large samples of referred and non-referred children and adolescents which were matched for age, sex and socio-economic status. The combined sample that was used for statistical analysis consists of 2,058 referred and 2,058 non-referred boys and girls between 4 and 18 years of age (mean age: referred boys= 10.9 years, non-referred boys = 10.9 years, referred girls=11.3 years, non-referred girls=11.1 years). Referral status was used as validity criterion. Statistical procedures included Odds Ratios, Total Predictive Values, ROC analyses and discriminant analyses. Results indicated that the discriminant validity of the German version of CBCL is comparable to the original English version. With the use of CBCL Total Problem Score as predictor (cut-off T > or = 60) 83.8% of children and adolescents could correctly be classified (sensitivity 83.6%, specificity 83.9%). Symptoms of the "Attention Problems Scale" show the highest discriminative power to distinguish between disturbed and undisturbed children and adolescents.

MeSH terms

  • Adolescent
  • Child
  • Child Behavior / psychology*
  • Child, Preschool
  • Discriminant Analysis
  • Female
  • Germany
  • Humans
  • Male
  • Odds Ratio
  • Personality Inventory
  • ROC Curve
  • Reproducibility of Results