This course is an introduction to linear structural equation modeling and its application to social and behavioral research. This course covers various data analytic techniques ranging from simple regression and path analysis to confirmatory factor analysis, multiple group models, longitudinal growth models, categorical outcomes, mixture modeling, and power analysis. For each topic, the algebraic and graphic model specification, estimation procedures, identification, goodness-of-fit criteria, and alternative model comparison approaches are discussed. The goals of this course are to develop an understanding of the conceptual and mathematical bases of structural equation modeling, to learn how to specify and estimate models using statistical software, and evaluate structural equation models in relation to alternative models using statistical and practical criteria.
Several software packages are discussed throughout this course. Primarily, Mplus, a commercial software, and lavaan (latent variable analysis), a package in R, are used to specify and estimate the structural equation models discussed. Additionally, we discuss the Lisrel (linear structural relations), a commercial software, because this program established a common notation for structural equation models that continues to be the dominant matrix notation used by researchers. R is available at http://cran.us.r-project.org, and the lavaan package can be installed by specifying install.packages(‘lavaan’). Please download the RStudio software (https://rstudio.com/), which provides a user-friendly front-end interface for R. A demonstration version of Mplus is available at http://www.statmodel.com/demo.shtml. Standard statistical analyses and matrix algebra are conducted using R.
Presentations & Handout
Lectures along with lecture notes, input and output scripts, and data are available on the Longitudinal Research Institute Webpage (https://longitudinalresearchinstitute.com/) and grouped by topic. The lecture notes, input and output scripts, and data are contained in a zip file associated with each topic
For most topics there is an associated activity folder. Review the handout for each activity, and then complete the activity, which often includes data analysis. Data for the activities are provided. Solutions to the activities are provided so you can check your work and understanding.