Multilevel Analysis

This practical introduction helps readers apply multilevel techniques to their research. Noted as an accessible introduction, the book also includes advanced extensions, making it useful as both an introduction and as a reference to students, researchers, and methodologists. Basic models and examples are discussed in non-technical terms with an emphasis on understanding the methodological and statistical issues involved in using these models. The estimation and interpretation of multilevel models is demonstrated using realistic examples from various disciplines. For example, readers will find data sets on stress in hospitals, GPA scores, survey responses, street safety, epilepsy, divorce, and sociometric scores, to name a few. The data sets are available on the website in SPSS, HLM, MLwiN, LISREL and/or Mplus files. Readers are introduced to both the multilevel regression model and multilevel structural models. Highlights of the second edition include: Two new chapters—one on multilevel models for ordinal and count data (Ch. 7) and another on multilevel survival analysis (Ch. 8). Thoroughly updated chapters on multilevel structural equation modeling that reflect the enormous technical progress of the last few years. The addition of some simpler examples to help the novice, whilst the more complex examples that combine more than one problem have been retained. A new section on multivariate meta-analysis (Ch. 11). Expanded discussions of covariance structures across time and analyzing longitudinal data where no trend is expected. Expanded chapter on the logistic model for dichotomous data and proportions with new estimation methods. An updated website at http://www.joophox.net/ with data sets for all the text examples and up-to-date screen shots and PowerPoint slides for instructors. Ideal for introductory courses on multilevel modeling and/or ones that introduce this topic in some detail taught in a variety of disciplines including: psychology, education, sociology, the health sciences, and business. The advanced extensions also make this a favorite resource for researchers and methodologists in these disciplines. A basic understanding of ANOVA and multiple regression is assumed. The section on multilevel structural equation models assumes a basic understanding of SEM.