Optimal use of multi-informant data on co-occurrence of internalizing and externalizing problems: the TRAILS study.

Strong between-informant discrepancies are found in ratings of (pre)adolescent problems and in co-occurrence rates between different domains of psychopathology. These discrepancies can be caused by differences in the context of measurement and the perspective of informants. The aim of this study was to develop a "Multi-Informant Co-occurrence" model (MIC), which takes into account these differences in context and perspective. In a population-based cohort of (pre)adolescents (n = 2230) from a longitudinal study in the north of the Netherlands, internalizing (INT) and externalizing (EXT) problems were rated by the (pre)adolescents themselves, their teachers, and their parents. As hypothesized Principal Component Analysis revealed four independent main components: Between-domain convergence was captured by a severity component (S), while between-domain discrepancy was captured by a direction component (D). Between-informant discrepancies were captured by a perspective (P) and a context (C) component. The use of this MIC-model will increase reliability and validity of measures of psychopathology and the four components each provide useful specific information.