Principles and Practice of Structural Equation Modeling

What distinguishes this reading route

The contents address nonparametric and parametric causal models, covariate selection, local estimation and piecewise SEM, followed by global estimation and testing. Measurement coverage distinguishes reflective, formative and composite indicators before CFA. SEM becomes a family of modeling arguments rather than a single software interface.

Model comparison and respecification also invite careful reading. A change made after examining data should be distinguished from an original theoretical commitment. That distinction supports more transparent methods and discussion sections.

Read with an assumption to defend

For each diagram, describe what it allows and excludes. A covariate, indirect path or correlated residual makes a modeling commitment. This is more informative than asking only how to improve fit.

Survey researchers should be particularly careful with causal wording. The book opens a route into causal assumptions; it does not license treating every estimated path as a causal effect. Design, timing and confounding remain study-specific concerns.

Exercise created by Lê Minh: draw two hypothetical diagrams for the same three variables. State what evidence could distinguish their stories and which assumptions require defense. Read specification and comparison sections to consider why choosing the better-fitting diagram may not settle the issue.

Audience and edition

This is the 2023 fifth edition, ISBN 9781462551910, distinct from the separately introduced 2011 third edition. It suits readers with basic statistics, doctoral researchers and critical reviewers. Its payoff is the ability to explain a modeling decision and its limits, a more durable skill than recalling another fit threshold.

Choose a companion reading

Continue with: Handbook of Structural Equation Modeling (Second edition · English); Principles and Practice of Structural Equation Modeling (Third edition · 2011 · English).

Source and edition