Random effects regression modeling of spatial and longitudinal data

点击翻译按钮获取中文摘要

It is well known that the poor health of adolescents is associated with unhealthy life-styles and participation in risky activities such as smoking, alcohol and drug use. Understanding when, why and how risk behaviours are initiated helps deliver effective interventions in schools. As these risk behaviours are usually group behaviours associated with underlying community characteristics, I am interested in the development of multivariate random effects models where random effects can be used to account for such community effects and the association among neighbouring communities. These models help understand the nature of regional disparity across Canada and identify factors which improve social outcomes and reduce inequalities. These findings will help government decision makers geographically target resource allocation, health promotion, education policy and regional planning in an attempt to remediate problem areas.