This project involves the study of a sample of children in Dominica. The project builds on long-term research on variations in corticosteroid stress response and socio-cultural variables. Using randomized scan sampling techniques as well as standard participant observation and interviews with family members, the project will assess the behavioral correlates to biochemical variation in hormonal assays related to stress in a sample of 126 children. This sort of research is important because the many studies of the causes of children's stress have not examined biochemical variables as they vary with micro-level differences in family composition and interaction styles. This project will generate data that will be useful in advancing our understanding of the causes of childhood stress.