Structural Equation Modeling with Mplus: Basic Concepts, Applications,and Programming

Barbara M. Byrne

Routledge, 2012, ix + 278 pages, £85.00/$135.00, hardcover (also available as softcover)

Table of Contents

  • Section I. Introduction

    • 1.
      Structural equation models: the basics
    • 2.
      Using the Mplus program
  • Section II. Single-Group Analyses

    • 3.
      Testing the factorial validity of a theoretical construct: first-order confirmatory factor analysis model
    • 4.
      Testing the factorial validity of scores from a measuring instrument: first-order confirmatory factor analysis model
    • 5.
      Testing the validity of scores from a measuring instrument: second-order confirmatory factor analysis model
    • 6.
      Testing the validity of a causal structure: full structural equation model
  • Section III. Multiple-Group Analyses

    • 7.
      Testing for the factorial equivalence of a measuring instrument: analysis of covariance structures
    • 8.
      Testing for the equivalence of latent factor means: analysis of mean and covariance structures
    • 9.
      Testing for the equivalence of a causal structure: analysis of covariance structures
  • Section IV. Other Important Topics

    • 10.
      Testing evidence of construct validity: the multitrait–multimethod model
    • 11.
      Testing change over time: the latent growth curve model
    • 12.
      Testing within- and between-level variability: the multilevel model

Readership: Students of behavioural and social sciences interested in learning the basics of structural equation modelling (SEM) with Mplus through real-world examples.

As a basis for this short review, I have a fresh experience of using this book and its datasets as the primary material of a basic SEM course. Some of my comments will also reflect the feedback I got from my students during the course.

In general, I was happy with this book. It is clearly written, it focuses on important topics that are built systematically, and it covers a wide selection of interesting examples from the author's own research. It also stresses several important aspects of the interplay between statistical modelling and substantial reasoning.

The book is targeted to non-mathematical readers, and hence it focuses on the applications of SEM. It does this very nicely, beginning from the part that covers the basic ideas of SEM and shows how to get started with the Mplus. After the intro, the chapters are like small studies starting from research questions, hypothesised models and data description, followed by the actual analyses and results that are carefully ‘walked through’.

Each chapter introduces a new aspect, a study design or a model, typically adding a bit more complexity compared to the preceding ones. This is a useful way of proceeding and supporting the learning. Many of the chapters also bring out some additional points of important topics like missing values, categorical or non-normal variables. It is very instructive to demonstrate these challenges with real-world examples.

Some minor problems have to be mentioned. There are not many formulas, but surprisingly many of those given (in order to help certain calculations) include errors, e.g., missing parentheses. At times some details of the modelling processes remain unclear, causing the conclusions to be a bit vague. In a few examples, the items of a certain measuring instrument have not been described properly due to copyright restrictions, which makes the analyses and interpretations quite technical and less interesting. The syntax files were not available on the web in spite of the promise on the back cover. However, all these problems can be overcome or even considered as ‘extra challenges’.

Overall, this book is an excellent resource for a beginner interested in SEM with Mplus.