Research--Arsenault et al

Previous studies have reported significant heritabilities for aggressive behavior in children and adolescents. Most of these studies have used data from only one source, such as ratings by parents. Because questions have been raised about potential biases in data obtained from parents and other kinds of informants, a team of researchers at the Institute of Psychiatry in London evaluated the contributions of four sources of data to tests of genetic effects on young children's antisocial behavior (Arsenault et al., 2003). The children were 2,232 5-year-old British twins who were participating in the Environmental Risk (E-risk) Longitudinal Twin Study. The assessment data included mothers' CBCL ratings, teachers' TRF ratings, ratings by examiners who observed the children in standardized situations, and children's self-reports obtained in puppet interviews where each child identified puppets whose behavior was most like the child's own behavior. Correlations between scores obtained from the different informants (mothers, teachers, observers, children) were similar to the low correlations found between these combinations of informants in meta-analyses of many studies. However, complex statistical analyses indicated that antisocial behavior agreed upon by all types of informants (i.e., antisocial behavior that was observed in multiple contexts) was highly heritable, yielding a heritability estimate of 82%. Furthermore, the low correlations among scores from different informants did not result from biases in the informants' ratings. Instead, the variations in antisocial behavior that were specific to reports by each type of informant were influenced by genetic factors. In other words, the low correlations between informants' reports resulted from the fact that each type of informant validly captured different, genetically influenced aspects of children's antisocial behavior. These different aspects of antisocial behavior that were detected by each type of informant were thus in addition to the genetically influenced aspects of antisocial behavior that were consistent across multiple informants and the contexts to which their ratings referred. Arsenault et al. concluded that "researchers studying children's behavior disorders should try to collect data from different sources," and that "using all the information from each rater to create composite scores will capture more meaningful variation than restricting composites to only the information agreed upon by all raters" (p. 844).

Reference: Arsenault, L., Moffitt, T.E., Caspi, A., Taylor, A., Rijsdijk, F.V., Jaffee, S.R., Ablow, J.C, & Measelle, J.R. (2003). Strong genetic effects on cross-situational antisocial behaviour among 5-year-old children according to mothers, teachers, examiner-observers, and twins' self-reports. Journal of Child Psychology and Psychiatry, 44, 832-848.