Emphasizing concepts and rationale over mathematical minutiae, this is the most widely used, complete, and accessible structural equation modeling (SEM) text. Continuing the tradition of using real data examples from a variety of disciplines, the significantly revised fourth edition incorporates recent developments such as Pearl's graphing theory and the structural causal model (SCM), measurement invariance, and more. Readers gain a comprehensive understanding of all phases of SEM, from data collection and screening to the interpretation and reporting of the results. Learning is enhanced by exercises with answers, rules to remember, and topic boxes. The companion website supplies data, syntax, and output for the book's examples--now including files for Amos, EQS, LISREL, Mplus, Stata, and R (lavaan).
New to This Edition
*Extensively revised to cover important new topics: Pearl's graphing theory and the SCM, causal inference frameworks, conditional process modeling, path models for longitudinal data, item response theory, and more.
*Chapters on best practices in all stages of SEM, measurement invariance in confirmatory factor analysis, and significance testing issues and bootstrapping.
*Expanded coverage of psychometrics.
*Additional computer tools: online files for all detailed examples, previously provided in EQS, LISREL, and Mplus, are now also given in Amos, Stata, and R (lavaan).
*Reorganized to cover the specification, identification, and analysis of observed variable models separately from latent variable models.
*Exercises with answers, plus end-of-chapter annotated lists of further reading.
*Real examples of troublesome data, demonstrating how to handle typical problems in analyses.
*Topic boxes on specialized issues, such as causes of nonpositive definite correlations.
*Boxed rules to remember.
*Website promoting a learn-by-doing approach, including syntax and data files for six widely used SEM computer tools.
"Kline is a master at explaining complex concepts in a very accessible manner. It is refreshing to see a new edition of an important book that truly is new, not simply redesigned. The fourth edition successfully incorporates recent developments in SEM and contemporary forms of causal reasoning and analysis, such as the SCM. Unlike most SEM texts, this book is notable for making a sophisticated, often-difficult statistical technique understandable to non-statisticians without watering down the material. Kline makes excellent use of relevant statistical theory without overwhelming the reader with algebraic matrices, proofs, formulas, and statistical notations. I recommend this book without reservation to researchers, instructors, and students in the social and behavioral sciences. It is far more than an introduction to SEM--in my opinion, it is a potential catalyst for reconsidering the statistical methods that researchers apply to better understand human action and interaction."--Chris L. S. Coryn, PhD, Director, Interdisciplinary PhD in Evaluation, Western Michigan University
"Too often, new editions of statistics books do not have substantive changes, but that is not the case here--Kline has made significant improvements to an already excellent book. Staying current is particularly necessary in SEM, where the theory has been developing rapidly in the last 10 years, yielding, for example, better estimation methods for categorical data and Bayesian methods. Helpful features include the topic boxes, which allow detailed discussion of particular topics without interfering with the overall flow of the text. I also like the exercises at the end of each chapter, which highlight the important parts of the chapter and provide crucial learning opportunities. Kline's use of the companion website to distribute real examples is excellent. After reading about the models and analyses, it is helpful--actually vital--to be able to practice running the models in various software packages."--Craig S. Wells, PhD, Department of Educational Policy, Research, and Administration, University of Massachusetts Amherst
"The best place to start for anyone who wants to learn the basics of SEM. The text emphasizes applied SEM content without relying on statistical formulas and the writing is clear and well organized, which is very helpful for students. I appreciate having exercises with answers that students can complete and check on their own. The examples are very helpful, and reflect the fact that real data are often troublesome. The website is easy to use and more extensive than for many other books."--Donna Harrington, PhD, University of Maryland School of Social Work
"The incorporation of Pearl's approach to causal inference is a major improvement in the fourth edition. This is the most useful introductory SEM book out there. I have recommended this book to colleagues for both personal and class use, and will continue to do so."--Richard K. Wagner, PhD, Distinguished Professor of Psychology, Florida State University; Associate Director, Florida Center for Reading Research
"This book is unique in that it treats structural equation models for what they are--carriers of causal assumptions and tools for causal inference. Gone are the inhibitions and trepidation that characterize most SEM texts in their treatments of causal inference. Overall, the book elevates SEM education to a new level of modernity and promises to usher in a renaissance for a field that pioneered causal analysis in the behavioral sciences."--Judea Pearl, PhD, Department of Computer Science, University of California, Los Angeles
"Perfectly addresses the needs of social scientists like me without formal training in mathematical statistics....Can be read by any graduate in psychology or even by keen undergraduates interested in exploring new vistas. Yet it will also constitute a surprisingly good read for experienced researchers in search of some refreshing insights in their favorite techniques....A real tour de force....Succeeds in reconciling comprehensiveness and comprehensibility."--The Psychologist (on the second edition)
"The greatest strength of this book is Kline's ability to present materials in an engaging, accessible manner. In nearly all situations, Kline is able to describe even the more complex material in practical, jargon-free terms....In this regard, this book is unparalleled, and I suspect that this strength alone will make this the book of choice for many who are eager to learn SEM but who do not possess extensive quantitative backgrounds...This book could be readily adapted to courses for students with a basic understanding of correlation and regression or as part of a course for more advanced students."--PsycCRITIQUES (on the second edition)
I. Concepts and Tools 1. Coming of Age Preparing to Learn SEM Definition of SEM Importance of Theory A Priori, but Not Exclusively Confirmatory Probabilistic Causation Observed Variables and Latent Variables Data Analyzed in SEM SEM Requires Large Samples Less Emphasis on Significance Testing SEM and Other Statistical Techniques SEM and Other Causal Inference Frameworks Myths about SEM Widespread Enthusiasm, but with a Cautionary Tale Family History Summary Learn More 2. Regression Fundamentals Bivariate Regression Multiple Regression Left-Out Variables Error Suppression Predictor Selection and Entry Partial and Part Correlation Observed versus Estimated Correlations Logistic Regression and Probit Regression Summary Learn More Exercises 3. Significance Testing and Bootstrapping Standard Errors Critical Ratios Power and Types of Null Hypotheses Significance Testing Controversy Confidence Intervals and Noncentral Test Distributions Bootstrapping Summary Learn More Exercises 4. Data Preparation and Psychometrics Review Forms of Input Data Positive Definiteness Extreme Collinearity Outliers Normality Transformations Relative Variances Missing Data Selecting Good Measures and Reporting about Them Score Reliability Score Validity Item Response Theory and Item Characteristic Curves Summary Learn More Exercises 5. Computer Tools Ease of Use, Not Suspension of Judgment Humana "Computer Interaction Tips for SEM Programming SEM Computer Tools Other Computer Resources for SEM Computer Tools for the SCM Summary Learn More II. Specification and Identification 6. Specification of Observed Variable (Path) Models Steps of SEM Model Diagram Symbols Causal Inference Specification Concepts Path Analysis Models Recursive and Nonrecursive Models Path Models for Longitudinal Data Summary Learn More Exercises Appendix 6.A. LISREL Notation for Path Models 7. Identification of Observed Variable (Path) Models General Requirements Unique Estimates Rule for Recursive Models Identification of Nonrecursive Models Models with Feedback Loops and All Possible Disturbance Correlations Graphical Rules for Other Types of Nonrecursive Models Respecification of Nonrecursive Models that are Not Identified A Healthy Perspective on Identification Empirical Underidentification Managing Identification Problems Path Analysis Research Example Summary Learn More Exercises Appendix 7.A. Evaluation of the Rank Condition 8. Graph Theory and the Structural Causal Model Introduction to Graph Theory Elementary Directed Graphs and Conditional Independences Implications for Regression Analysis d-Separation Basis Set Causal Directed Graphs Testable Implications Graphical Identification Criteria Instrumental Variables Causal Mediation Summary Learn More Exercises Appendix 8.A. Locating Conditional Independences in Directed Cyclic Graphs Appendix 8.B. Counterfactual Definitions of Direct and Indirect Effects 9. Specification and Identification of Confirmatory Factor Analysis Models Latent Variables in CFA Factor Analysis Characteristics of EFA Models Characteristics of CFA Models Other CFA Specification Issues Identification of CFA Models Rules for Standard CFA Models Rules for Nonstandard CFA Models Empirical Underidentification in CFA CFA Research Example Appendix 9.A. LISREL Notation for CFA Models 10. Specification and Identification of Structural Regression Models Causal Inference with Latent Variables Types of SR Models Single Indicators Identification of SR Models Exploratory SEM SR Model Research Examples Summary Learn More Exercises Appendix 10.A. LISREL Notation for SR Models III. Analysis 11. Estimation and Local Fit Testing Types of Estimators Causal Effects in Path Analysis Single-Equation Methods Simultaneous Methods Maximum Likelihood Estimation Detailed Example Fitting Models to Correlation Matrices Alternative Estimators A Healthy Perspective on Estimation Summary Lean More Exercises Appendix 11.A. Start Value Suggestions for Structural Models 12. Global Fit Testing State of Practice, State of Mind A Healthy Perspective on Global Fit Statistics Model Test Statistics Approximate Fit Indexes Recommended Approach to Fit Evaluation Model Chi-Square RMSEA CFI SRMR Tips for Inspecting Residuals Global Fit Statistics for the Detailed Example Testing Hierarchical Models Comparing Nonhierarchical Models Power Analysis Equivalent and Near-Equivalent Models Summary Learn More Exercises Appendix 12.A. Model Chi-Squares Printed by LISREL 13. Analysis of Confirmatory Factor Analysis Models Fallacies about Factor or Indicator Labels Estimation of CFA Models Detailed Example Respecification of CFA Models Special Topics and Tests Equivalent CFA Models Special CFA Models Analyzing Likert-Scale Items as Indicators Item Response Theory as an Alternative to CFA Summary Learn More Exercises Appendix 13.A. Start Value Suggestions for Measurement Models Appendix 13.B. Constraint Interaction in CFA Models 14. Analysis of Structural Regression Models Two-Step Modeling Four-Step Modeling Interpretation of Parameter Estimates and Problems Detailed Example Equivalent Structural Regression Models Single Indicators in a Nonrecursive Model Analyzing Formative Measurement Models in SEM Summary Learn More Exercises Appendix 14.A. Constraint Interaction in SR Models Appendix 14.B. Effect Decomposition in Nonrecursive Models and the Equilibrium Assumption Appendix 14.C. Corrected Proportions of Explained Variance for Nonrecursive Models IV. Advanced Techniques and Best Practices 15. Mean Structures and Latent Growth Models Logic of Mean Structures Identification of Mean Structures Estimation of Mean Structures Latent Growth Models Detailed Example Comparison with a Polynomial Growth Model Extensions of Latent Growth Models Summary Learn More Exercises 16. Multiple-Samples Analysis and Measurement Invariance Rationale of Multiple-Samples SEM Measurement Invariance Testing Strategy and Related Issues Example with Continuous Indicators Example with Ordinal Indicators Structural Invariance Alternative Statistical Techniques Summary Learn More Exercises Appendix 16.A. Welcha "James Test 17. Interaction Effects and Multilevel Structural Equation Modeling Interactive Effects of Observed Variables Interactive Effects in Path Analysis Conditional Process Modeling Causal Mediation Analysis Interactive Effects of Latent Variables Multilevel Modeling and SEM Summary Exercises Learn More 18. Best Practices in Structural Equation Modeling Resources Specification Identification Measures Sample and Data Estimation Respecification Tabulation Interpretation Avoid Confirmation Bias Bottom Lines and Statistical Beauty Summary Learn More Suggested Answers to Exercises References Author Index Subject Index About the Author