ABSTRACT
The purposes of this chapter are to provide a guide to group treatment researchers for research design that is informed by multilevel model (MLM) of nested change data and to provide a technical primer for developing and using three-level MLM equations specific to group treatment research contexts. The defining design element of group treatment research is that clients are organized into groups, and so the client data are nested within groups. Nesting refers to data with a multilevel or hierarchical structure wherein units are clustered into a hierarchy. A common method for assessing longitudinal data is by the repeated measurement random analysis of variance (rANOVA) model in which mean differences across time are compared. A restriction of rANOVA is the requirement of having complete data for each individual at each time point. The primary issue that arises in research design with hierarchically nested data is the unit of analysis problem.
