The study of teachers as decision-makers is a recent paradigm in educational research. This decision-making approach considers the teacher as an active professional engaged in various decision-making activities prior to and during instruction. This paper describes a naturalistic research design employed in a recent dissertation that investigated factors affecting teachers' diagnostic decisions. The objective of the paper is to present this research design as an effective alternative or complement to quasi-experimental designs in the study of factors affecting teachers' decision-making. The research design employed the in-depth and interpretative focus of a multiple case study approach and methodological triangulation within naturalistic settings. Specifically, the basic design of the study was a fixed multiple case study investigation of two experimental groups and no control groups, utilizing repeated measures within a multi-method approach to identify differences among groups rather than any effects of treatment. The multiple case studies employed in this study approach were extremely effective in providing a rich source of qualitative and quantitative data. In addition, the safeguards built into the design ensured the trustworthiness of this data. (Author)