Multilevel Modeling Techniques: MLM and SEM Approaches Using Mplus
=An Introduction to Multilevel Modeling Techniques: MLM and SEM Approaches Using Mplus,3rd EditionAuthor(s): Ronald H. Heck, Scott L. Thomas

This coursebook provides an applied approach for utilizing multilevel modeling techniques within these fields and disciplines. Our intent is to develop a basic rationale behind the use of these techniques and to provide an introduction to the process of developing, testing, and interpreting the results of models that facilitate the investigation of hierarchical data structures. Hierarchical (or nested) data structures are defined by the nesting of a lower-level unit of analysis in a higher-level grouping that may itself constitute a separate unit of analysis. Individuals, for example, may be nested in various types of higher-order groupings such as employees clustered within departments and within companies, students clustered within classrooms and within schools, patients clustered within nursing units within hospitals, and repeated measures nested within individuals who may be randomly assigned to various experimental and treatment groups. Single-level analyses of hierarchical data would not be appropriate in most situations because clustering suggests that individuals within groups may be more similar to each other than to individuals clustered within other groups. Treating individuals as if they were separate from their social groupings therefore introduces potential biases in the proper analysis of hierarchical data structures. Along the way in our presentation of multilevel modeling, we provide numerous examples of crosssectional and longitudinal hierarchical data structures with outcome variables scaled at a range of measurement levels including nominal, dichotomous, ordinal, count, and interval/ratio. 。。。。。。。。。。。