Approaches to Understanding Structural Models: Models of Relationships Between Variables, Occasions, and People

by Phillip Karl Wood

This is a book format for lecture notes I developed for my classes in structural equation modeling. It describes how to specify, evaluate, and compare Structural Equation Models (SEMs). Central ideas of SEM's are presented via scatterplots, path diagrams, Equations, vector diagrams, and matrix algebra. It emphasizes that one not only specify one model, but think critically about other counterarguments a reasonable skeptic might offer for a proposed model. Numerous examples are present which illustrate confirmatory factor models, various growth curve models, multi-group models, measurment models, and exploratory factor models. Assumption checking, estimation, and rotation techniques are also discussed. Examples of analyses done in Amos, Lavaan, Mplus, and Onyx are presented. An appendix of matrix algebra ideas relevant to SEM specification is presented as well. I expect the material covered exceeds what a one semester introduction to SEM could cover, but readers can select chapters relevant to their particular research needs.I expect I will flesh out the exploratory factor analysis discussion in a subsequent version and plan to include chapters on Bayesian estimation and model comparison as well, but the present manuscript would probably be a good starting point for many.

309 pages, Paperback

Published January 12, 2019