Multilevel Structural Equation Modeling by Bruno Castanho Silva, Constantin Manuel Bosancianu, and Levente Littvay serves as a minimally technical overview of multilevel structural equation modeling (MSEM) for applied researchers and advanced graduate students in the social sciences. As the first book of its kind, this title is an accessible, hands-on introduction for beginners of the topic. The authors predict a growth in this area, fueled by both data availability and also the availability of new and improved software to run these models. The applied approach, combined with a graphical presentation style and minimal reliance on complex matrix algebra guarantee that this volume will be useful to social science graduate students wanting to utilize such models.

"This accessible volume presents both the mechanics of structural equation modeling (SEM) and specific SEM strategies and applications. The editor, along with an international group of contributors, and editorial advisory board are leading methodologists who have organized the book to move from simpler material to more statistically complex modeling approaches. Sections cover the foundations of SEM; statistical underpinnings, from assumptions to model modifications; steps in implementation, from data preparation through writing the SEM report; and basic and advanced applications, including new and emerging topics in SEM. Each chapter provides conceptually oriented descriptions, fully explicated analyses, and engaging examples that reveal modeling possibilities for use with readers' data. Many of the chapters also include access to data and syntax files at the companion website, allowing readers to try their hands at reproducing the authors' results"--

Univariate and multivariate multilevel models are used to understand how to design studies and analyze data in this comprehensive text distinguished by its variety of applications from the educational, behavioral, and social sciences. Basic and advanced models are developed from the multilevel regression (MLM) and latent variable (SEM) traditions within one unified analytic framework for investigating hierarchical data. The authors provide examples using each modeling approach and also explore situations where alternative approaches may be more appropriate, given the research goals. Numerous examples and exercises allow readers to test their understanding of the techniques presented. Changes to the new edition include: -The use of Mplus 7.2 for running the analyses including the input and data files at www.routledge.com/9781848725522. -Expanded discussion of MLM and SEM model-building that outlines the steps taken in the process, the relevant Mplus syntax, and tips on how to evaluate the models. -Expanded pedagogical program now with chapter objectives, boldfaced key terms, a glossary, and more tables and graphs to help students better understand key concepts and techniques. -Numerous, varied examples developed throughout which make this book appropriate for use in education, psychology, business, sociology, and the health sciences. -Expanded coverage of missing data problems in MLM using ML estimation and multiple imputation to provide currently-accepted solutions (Ch. 10). -New chapter on three-level univariate and multilevel multivariate MLM models provides greater options for investigating more complex theoretical relationships(Ch.4). -New chapter on MLM and SEM models with categorical outcomes facilitates the specification of multilevel models with observed and latent outcomes (Ch.8). -New chapter on multilevel and longitudinal mixture models provides readers with options for identifying emergent groups in hierarchical data (Ch.9). -New chapter on the utilization of sample weights, power analysis, and missing data provides guidance on technical issues of increasing concern for research publication (Ch.10). Ideal as a text for graduate courses on multilevel, longitudinal, latent variable modeling, multivariate statistics, or advanced quantitative techniques taught in psychology, business, education, health, and sociology, this book’s practical approach also appeals to researchers. Recommended prerequisites are introductory univariate and multivariate statistics.

By focusing on how structural equation modeling (SEM) can illuminate what we know abou social and behavioral phenomena rather than on the "ins and outs" of the various SEM programs (eg., AMOS, LISREL, EQS), David Kaplan demonstrates to the reader the utility of SEM, its limitations, and the statistical underpinnings of the method. The book begins with an explanation of confirmatory factor analysis and path analysis, and how these two techniques can be combined for SEM.