Features

• Presents the history and theory of QLS regression
• Demonstrates how QLS can be used to extend the traditional generalized estimating equation (GEE) approach
• Considers several important data structures, including familial, multi-level, and unequally spaced longitudinal data in addition to the standard structures
• Provides methods for sample size calculation
• Includes approaches for the selection of a working structure and for the assessment of the goodness of fit of the model
• Covers the theoretical underpinnings of the methods in addition to copious detailed examples
• Compares QLS with popular alternate approaches
• Uses Stata for model construction and analysis, with corresponding R, SAS, and MATLAB code

Summary

Drawing on the authors’ substantial expertise in modeling longitudinal and clustered data, this book presents a comprehensive treatment of quasi-least squares (QLS) regression—a computational approach for the estimation of correlation parameters within the framework of generalized estimating equations (GEE). The authors present an overview and detailed evaluation of QLS methodology, demonstrating the advantages of QLS in comparison with alternative methods. A fully worked out example is provided that leads readers from the planning stages of a study, including sample size considerations, through model construction and interpretation. Special focus is given to goodness-of-fit analysis and strategies on selecting the appropriate working correlation structure. The text includes additional examples throughout to demonstrate each topic of discussion and uses Stata for the majority of examples, along with corresponding R, SAS, and MATLAB^® code.