Data Analysis with Mplus
Guilford Press – 2013 – 305 pages
A practical introduction to using Mplus for the analysis of multivariate data, this volume provides step-by-step guidance, complete with real data examples, numerous screen shots, and output excerpts. The author shows how to prepare a data set for import in Mplus using SPSS. He explains how to specify different types of models in Mplus syntax and address typical caveats—for example, assessing measurement invariance in longitudinal SEMs. Coverage includes path and factor analytic models as well as mediational, longitudinal, multilevel, and latent class models. Specific programming tips and solution strategies are presented in boxes in each chapter. The companion website features data sets, annotated syntax files, and output for all of the examples. Of special utility to instructors and students, many of the examples can be run with the free demo version of Mplus.
"Mplus is arguably the most flexible commercially available software program for SEM and all of its special cases. Geiser has provided an admirable service to the community of researchers who use Mplus with this highly readable book. The book is an indispensable companion to more advanced SEM texts and is certainly an important supplementary text for graduate courses on SEM." - David Kaplan, PhD, , University of Wisconsin-Madison, USA
"More and more researchers all over the world are using Mplus. I know of no other book that provides such a truly helpful tutorial on everything from the very first steps to how to run complicated SEM models like latent growth models. Beginners will very much appreciate how much attention the author pays to the basics. Many easy-to-make mistakes can be prevented by keeping this book within arm's reach. It is perfect for researchers at any career stage seeking an accessible, informative introduction to analyzing data with Mplus." - Rens van de Schoot, PhD, Utrecht University, The Netherlands
Part I: Data Management in SPSS. Coding Missing Values. Exporting an ASCII Data File for Mplus. Part II: Reading Data into Mplus. Importing and Analyzing Individual Data (Raw Data). Basic Structure of the Mplus Syntax and Basic Analysis. Mplus Output for Basic Analysis. Importing and Analyzing Summary Data (Covariance or Correlation Matrices). Part III: Linear Structural Equation Models. What are Linear SEMs? Simple Linear Regression Analysis with Manifest Variables. Latent Regression Analysis. Confirmatory Factor Analysis. First-Order CFA. Second-Order CFA. Path Models and Mediator Analysis. Introduction and Manifest Path Analysis. Manifest Path Analysis in Mplus. Latent Path Analysis. Latent Path Analysis in Mplus. Part IV: Structural Equation Models for Measuring Variability and Change. Latent State Analysis. LS versus LST Models. Analysis of LS Models in Mplus. Modeling Indicator-Specific Effects. Testing for Measurement Invariance across Time. LST Analysis. Autoregressive Models. Manifest Autoregressive Models. Latent Autoregressive Models. 4 Latent Change Models. Latent Growth Curve Models. First-Order LGCMs. Second-Order LGCMs. Part V: Multilevel Regression Analysis. Introduction to Multilevel Analysis. Specification of Multilevel Models in Mplus. Option two level basic. Random Intercept Models. Null Model (Intercept-Only Model). One-Way Random Effects of ANCOVA. Means-as-Outcomes Model. Random Intercept and Slope Models. Random Coefficient Regression Analysis. Intercepts-and-Slopes-as-Outcomes Model. Part VI: Latent Class Analysis. Introduction to Latent Class Analysis. Specification of LCA Models in Mplus. Model Fit Assessment and Model Comparisons. Absolute Model Fit. Relative Model Fit. Interpretability. Appendix A: Summary of Key Mplus Commands Discussed in This Book. Appendix B: Common Mistakes in the Mplus Input Setup and Troubleshooting. Appendix C: Further Readings.
Christian Geiser, PhD, is Assistant Professor in the Department of Psychology at Utah State University. His methodological research interests include the development, evaluation, and application of latent variable psychometric models for longitudinal and multimethod data. His substantive research focuses on how individual