An Analysis of the Relationship Between Parent and Adolescent Marijuana Use Via Generalized Estimating Equation Methodology.

Although longitudinal designs are uniquely suited to the study of individual change over time, standard multivariate statistical methods often perform poorly when faced with common characteristics of longitudinal data such as correlated observations and missing observations. The purpose of the present study was to demonstrate the utility of a generalized estimating equations approach to the analysis of longitudinal data, while investigating the role of parent marijuana use on lagged changes in the marijuana use of their adolescent children over a 6- year period. In addition, the existence of significant trends in use was examined. Findings indicated that, over time, parental use of marijuana was positively related to their adolescent's use of the same substance, although there were significant differences in the strength of the relationship across gender and age groups. Results are discussed with reference to the importance and utility of the generalized estimating equations approach to the analysis of longitudinal data.