The research being proposed will employ two data bases to study the impact of various components of family environment (generally corresponding to indivisible human capital inputs and divisible social capital inputs) on several different dimensions of family well-being. Two interrelated studies are proposed. This first study is an individual research project that will employ an underutilized sibling data base from the High School and Beyond study to examine family effects on elements of child well-being measured during and shortly after the high school years (late adolescence). The individual project is sufficiently detailed that it can also form the basis for group research if so determined. The second study is a group project that will employ sibling data from the National Longitudinal Study of Youth to examine the impact of family environment on child well-being during early to middle childhood among children born to adolescent mothers. Both projects will make use of LISREL (or LISREL-like models) to estimate models of sibling resemblance in child well-being. These models provide estimates of both between and within family regression slopes, adjust for measurement error and allow comparison of important sub-groups. With respect to the individual research project, several specific aims will be pursued: (l) Specific aim #l seeks to replicate prior research on the potentially biasing effect of family environment when occupational status and earnings are regressed on schooling. (2) Specific aim #2 constitutes a search for a better specification of family environment using the richness of the data in the HSB sibling file, (3) Specific aim #3 seeks to test a causal model of educational attainment using sibling data. In specifying this model, it will be possible to test for the presence of family bias in a number of intermediate outcomes such as grades. (4) Specific aim #4 seeks to elaborate the role of family environment in the development of life course and educational expectations and values in children.
