"Generalized Linear Models For Categorical And Continuous Limited Dependent Variables | Statistics

Generalized Linear Models for Categorical and Continuous Limited Dependent Variables is designed for graduate students and researchers in the behavioral social health and medical sciences. It incorporates examples of truncated counts censored continuous variables and doubly bounded continuous variables such as percentages. The book provides broad but unified coverage and the authors integrate the concepts and ideas shared across models and types of data especially regarding conceptual links between discrete and continuous limited dependent variables. The authors argue that these dependent variables are if anything more common throughout the human sciences than the kind that suit linear regression. They cover special cases or extensions of models estimation methods model diagnostics and of course software. They also discuss bounded continuous variables boundary-inflated models and methods for modeling heteroscedasticity. Wherever possible the authors have illustrated concepts models and techniques with real or realistic datasets and demonstrations in R and Stata and each chapter includes several exercises at the end. The illustrations and exercises help readers build conceptual understanding and fluency in using these techniques. At several points the authors bring together material that has been previously scattered across the literature in journal articles software package documentation files and blogs. These features help students learn to choose the appropriate models for their purpose.|Generalized Linear Models for Categorical and Continuous Limited Dependent Variables | Statistics