Over the course of last summer, I developed a strong interest in hierarchical linear modeling in a collaborative work with Dr. Wes Snyder (Assistant Vice President for Research & Director of the Office of International Programs) as part of the Gates project to enhance technology for educational leaders. The project is funded by the Bill and Melinda Gates Foundation.

Hierarchical linear models (HLM) provide a conceptual framework and a flexible set of analytic tools to study a variety of social, political, and developmental (Biological) processes. HLM incorporate data from multiple levels in an attempt to determine the impact of individual and grouping factors upon some individual level outcomes. For example, student achievement may be a function of student level characteristics (e.g., IQ, study habits), classroom level factors (e.g., instruction style, textbook), school level factors (e.g., wealth), and so on. HLMs, or multilevel models, can incorporate such factors in a manner better than ordinary least squares since HLMs take into account error structures at each level. In Biology, animal and human studies of inheritance deal with a natural hierarchy where offspring are grouped within families. Hierarchy is usually referred to the fact that these problems consist of units grouped at different levels. Thus offspring may be the level 1 units in a 2-level structure where the level 2 units are the families: students may be level 1 units clustered within schools that are the level 2 units.

In this talk I will give an overview of HLM, consider the formulation of statistical models in educational applications, give several examples of the 2-level and 3-level structures, and explain the procedure of solving them using one of the available software.