The first part of a three-volume study, this report presents and justifies a research design for investigation of the relationship between the housing environment and the range of child development and family measures. The recommended design is nonexperimental in nature and focuses on comparisons of residents in selected housing programs (publicly supported residential housing complexes). In chapter 1 the results of a computer simulation of the entire investigation are used to calculate optimal decision rules for the conduct of the study. Basic problems of inference that are associated with investigation of the interaction of human subjects with their environment are reviewed in the context of independent variables in chapter 2. Chapters 3 and 4 deal with nonexperimental and experimental approaches respectively, with major emphasis given to the description and documentation of the recommended nonexperimental approach. The final chapter presents recommended instruments for the measurement of the cognitive, socioemotional, and physical development in subject children, plus supplementary measures of family well-being. Three appendices provide descriptions of the instruments and documentation of the simulation study. (Author/ND)