The definitive one-stop resource on structural equation modeling (SEM) from leading methodologists is now in a significantly revised second edition. Twenty-three new chapters cover model selection, bifactor models, item parceling, multitraitmultimethod models, exploratory SEM, mixture models, SEM with small samples, and more. The book moves from fundamental SEM topics (causality, visualization, assumptions, estimation, model fit, and managing missing data); to major model types focused on unobserved causes of covariance between observed variables; to more complex, specialized applications. Each chapter provides conceptually oriented descriptions, fully explicated analyses, and engaging examples that reveal modeling possibilities for use with the reader's data. The expanded companion website presents full data sets, code, and output for many of the chapters, as well as bonus selected chapters from the prior edition. New to This Edition *Chapters on additional topics not mentioned above: SEM-based meta-analysis, dynamic SEM, machine-learning approaches, and more. *Chapters include computer code associated with example analyses (in Mplus and/or the R package lavaan), along with written descriptions of results. *60% new material reflects a decade's worth of developments in the mechanics and application of SEM. *Many new contributors and fully rewritten chapters.
About the Author
Rick H. Hoyle, PhD, is Professor of Psychology and Neuroscience and Director of the Center for the Study of Adolescent Risk and Resilience at Duke University. He is a Fellow of the Association for Psychological Science, the American Psychological Association (Divisions 1, 5, 8, and 9), and the Society for Experimental Social Psychology. Dr. Hoyle has written extensively on structural equation modeling and other statistical and methodological strategies for the study of complex social and behavioral processes.