Content area
Full Text
Hans-Christoph Steinhausen1 and Christa Winkler Metzke2 Received October 22, 1996; accepted November 12, 1997
The Youth Self Report (YSR) was used in a Swiss epidemiological study with 1093 subjects aged 10-17 years. Internal consistency was good for the secondorder factors "Internalizing Problems, " "Externalizing Problems, " and the "Total Problem Score," whereas it was less satisfactory for almost all syndrome scales. In general internal consistency coefficients were slightly lower in the Swiss sample than in the U.S. sample. Correlations between the syndrome scales resulted in good replications of the original findings as obtained in the U.S. normative sample. Effect analyses showed that sex is more important than age and nationality (indigenous vs. immigrant subjects). However, all effects were small. Correspondence between YSR and Child Behavior Checklist scores showed that agreement between adolescents' and parents' reports is relatively small.
INTRODUCTION
Recent epidemiological research in mental health of children and adolescents has greatly profited from the development of standardized scales for the assessment of behavioral and emotional problems. These scales are theoretically based on an empirically based taxonomy in which individual cases are linked to taxomic groupings using operational definitions (Achenbach, 1993). In a questionnaire format these scales allow screening of large samples with relatively low costs. Typically, epidemiological studies in this field follow a two-stage approach with screening in a first phase and interviewing smaller subsamples of screen-positive subjects and controls in a second phase. Objective, reliable, and valid instruments are required for both stages.
Besides the quality of the instruments, another important issue in epidemiological studies of child and adolescent psychopathology is the informant. The more recent family of studies performed in the 1980s has typically tried to collect information from various sources, i.e., the parents, the teachers, and the child (Anderson Williams et al., 1987; Bird et al., 1988; Boyle et al., 1987; McGee et al., 1990; Verhulst et al., 1985). Preferably, the information collected from these different informants should have a similar format in order to allow for comparisons across sources of information. Cross-informant syndrome constructs (Achenbach, 1993) serve these purposes. However, so far, there is no easy solution available for handling the discrepant information on adolescent problem behavior as perceived by the parents and the youths themselves. In the recent...