Regression models for categorical dependent variables using Stata /
J. Scott Long, Departments of Sociology and Statistics, Indiana University, Bloomington, Indiana; Jeremy Freese, Department of Sociology and Institute for Policy Research, Northwestern University, Evanston, Illinois.
- Third edition.
- College Station, Texas : Stata Press Publication, StataCorp LP, 2014.
- xxiii, 589 pages : illustrations ; 24 cm
Includes bibliographical references (pages 561-568) and indexes.
List of figures -- Preface -- I. General information. Introduction -- Introduction to Stata -- Estimation, testing, and fit -- Methods of interpretation -- II. Models for specific kinds of outcomes. Models for binary outcomes : estimation, testing, and fit -- Models for binary outcomes : interpretation -- Models for ordinal outcomes -- Models for nominal outcomes -- Models for count outcomes.
After reviewing the linear regression model and introducing maximum likelihood estimation, Long extends the binary logit and probit models, presents multinomial and conditioned logit models and describes models for sample selection bias. "The goal of Regression Models for Categorical Dependent Variables Using Stata, Third Edition is to make it easier to carry out the computations necessary to fully interpret regression models for categorical outcomes by using Stata's margins command. Because the models are nonlinear, they are more complex to interpret. Most software packages that fit these models do not provide options that make it simple to compute the quantities useful for interpretation. In this book, the authors briefly describe the statistical issues involved in interpretation, and then they show how you can use Stata to perform these computations."--Back cover.
9781597181112
Stata. Stata.
Regression analysis. Social sciences--Statistical methods--Data processing. Regression analysis. Social sciences--Statistical methods--Data processing.