[Reliability and factorial validity of the Child Behavior Checklist--an analysis of a clinical and field sample].

The purpose of this study was to assess the reliability (internal consistency), the correlations of the scale scores and the factorial validity of the German version of the Child Behavior Checklist (CBCL). The analyses were performed in a sample of patients at two child and adolescent psychiatric facilities (N = 1653) and in a community sample (N = 1622). The analyses were based on the 1991 American version of the CBCL. The internal consistency of most of the scales was good to excellent (rtt > .80). The reliability of the scales "Withdrawn", "Somatic complaints" and "Social Problems" was not completely satisfactory (rtt > .70). The reliability of the syndrome scales "Thought Problems" and "Sex Problems" and the Competence Scales was unsatisfactory (rtt < .70). The internal consistency of the syndrome scales in the German and American samples was very similar. The correlations between the scales were predominantly in the lower range, with some in a medium range (rtt > .50). Factor analyses in the clinical sample confirmed that the syndrome scales had been well-designed. With the exception of the scales "Social Problems" and "Withdrawn", where the items have loadings on a common factor, the scales proved to be factorially valid.