Studies of teacher development face the issue of providing a valid and generalizable analysis reflecting the influence of differences in people and contexts and their interactions. To adopt a study designed upon a single set of data, whether extensive or intensive, is limiting--both types of data are needed. Effective research provides for a wide set of data gathering instruments to be used, and the construction of categories and constructs that are validated across individual participants and the populations from which they are drawn. Theories and interventions based upon such research will have more generalizability than those developed from a series of case study research projects. A study of the characteristics and processes of teacher development in the management domain, based on data from student teachers, beginning teachers, and expert classroom managers, illustrates use of this research approach. The study shows how cohort and case study components can be included in a study of teacher development in ways which allow the generalizability of the insights and findings to be studied with cross-sectional and longitudinal cohort data. The study resulted in identification of five management approaches for classifying belief, advice, and self-description responses and resulted in construction of five management schema types that trace development changes. (JDD)