ABSTRACT
The flexibility of the SEM approach makes it an ideal modeling strategy for researchers to use in examining two-level cross-sectional or longitudinal models where measurement error is important to include. The examples included in this chapter provide an introduction to various ways in which SEM can be used for conducting multilevel modeling-yielding greater conceptual clarity in defining relationships among variables and greater precision of the estimates where there are clustering features present. Their appropriate use, however, depends on guidance from strong theory to explicate relationships and the collection of quality data. As the techniques continue to become more readily available, their use in a variety of research fields should expand rapidly.
