ABSTRACT
Because of the inherently hierarchical nature of many learning environments, data collected in these environments are nested in structure. More specifically, students work in dyads or groups. These groups are nested in classrooms, classrooms in schools, and schools in local cultures and school districts. Design researchers working in field settings build theory and design products to support learning in such environments. Consequently, design researchers must deal constantly with data structures of the type described here. Although much design research is qualitative in nature, this chapter takes a quantitative perspective and describes how quantitative researchers have begun to deal with nested data structures and the complexities of building theory and drawing inferences when data have this nested structure. In this chapter, the hierarchical linear model (HLM) is described and it is shown how this model provides a conceptual and statistical mechanism for investigating simultaneously how phenomena at different levels interact with each other. In so doing, the aggregation concerns raised by Saari (this volume) are addressed.
