The dual trajectory approach: detecting developmental behavioural overlaps in longitudinal and intergenerational research

Bentrup Christina

Abstract

Prospective longitudinal study designs are often referred to as the ideal way to explain causal relationships. Nevertheless, the evaluation of such data often leads to a reduction of empirical information. This is done by eliminating the factor time by forming sum indices or by analysing correlations over only two points in time. Newer methods of longitudinal research resort to group-based trajectory models which enables the analyses of inter- and intraindividual changes over time. However, less attention has been paid to a very useful extension of this method: the dual trajectory approach which provides the opportunity to combine two separate trajectory models in one model. In this way, the relationship between two sequences of behaviours can be estimated simultaneously and checked for overlaps. This approach is introduced, and an application based on prospective panel data from the German Crime in the modern City study (CrimoC) is carried out for the intergenerational relationship between physical violent parenting and juvenile delinquency.