Multilevel Regression Models (3) Covers multilevel regression models for the analysis of nested or hierarchical data, including both contextual and longitudinal applications.
SOC 578 Multilevel Regression Models (3)
This course is devoted to statistical models for regression analysis of multilevel data. Multilevel data arise when cases are sampled at two or more levels, with each lower level subsumed within the next higher, such as residents within neighborhoods within cities or individuals within families. Such data almost always violate the independence assumption of ordinary least squares regression, and in recent years a wealth of more appropriate techniques have become available. These methods bring the full flexibility of multiple regression analysis to the analysis of multilevel data, enabling scholars to address a broad range of research questions. This course thoroughly covers the basic multilevel regression model and also devotes considerable time to more advanced topics such as analysis of data with three or more levels, multilevel analysis of discrete dependent variables, and latent variables. Students will study examples in a broad range of substantive domains, with special attention to the unique research questions to which these methods give access. This is a course in the application of statistics to social science research, not a theoretical statistics course. Therefore the course will not include derivations and proofs, but rather the mathematics covered will be in the service of defining statistical models that correspond to useful research questions. The emphasis will be on understanding how to use these methods to do good research and on learning to interpret the results they provide. Several class sessions will be held in computer laboratories in order to train students in the use of statistical software that implements these methods.
Note : Class size, frequency of offering, and evaluation methods will vary by location and instructor. For these details check the specific course syllabus.