1S Structural Equation Modelling with MPLUS
Tom Scotto, University of Essex
9 – 20 July (two week course / 35 hours)
Detailed Course Outline [PDF]
THIS COURSE IS NOW FULLY BOOKED AND WE ARE OPERATING A WAITING LIST
Course Content
Structural Equation Modelling (SEM) is a framework that allows researchers to incorporate both measurement (factor) and path (causal) models into a single framework, and perform rigorous hypothesis testing of theoretically inspired models. Since its substantive origins in the late 1960s and early 1970s, SEM or “LISREL” modelling has come in and out of fashion in the applied social sciences. Today, it is undergoing a significant resurgence in part because the modelling framework has shown to be flexible enough to handle both cross-sectional and longitudinal designs, multi-level models, and unobserved heterogeneity. Estimators used to analyze dichotomous, ordinal, and censored data have also been developed to relax early (unrealistic) assumptions that the observed variables to be analyzed are normally distributed.
This course starts with the basics of Structural Equation Modelling and teaches students how to estimate SEMs using the flexible program MPLUS. No prior experience with SEM is assumed. The beginning of the first week covers the symbolic language and notation of SEM, traditional Exploratory Factor Analysis, and non-recursive path analysis using observed variables. The second half of the first week covers Confirmatory Factor Analysis (CFA), Model Identification and Fit, Model Constraints, and Incorporating CFA and Path Models into a Full SEM. The second week begins with lectures on Multiple Group Analysis, Multi-trait/Multi-Method Analyses, Model Modification, Recursive Models, and obtaining Indirect and Direct Effects from full blown SEMs. The course concludes with a discussion of advanced topics such as autoregressive models, latent growth curve modelling, and latent class analysis.
Course Objectives
The course objectives are to make students comfortable with the terminology, logic, use and abuse of the SEM framework. The early emphasis on the language of SEM will allow students not only to learn the basic material, but should make them comfortable reading more technical articles after completion of the course. Understanding the “language” of SEM will also allow them to keep up with the latest developments in the field. By the end of the course, students will not only understand the logic behind measurement and path modelling, they will also have proficiency in estimating such models using the MPLUS syntax based software package.
Course Prerequisites
Incoming students are assumed to have familiarity with regression analysis, an idea of why they may want to utilize SEM, and a willingness to learn basic matrix notation. Students should have experience with a statistical software package that can be used for cleaning data (SPSS, Stata, etc.). Students are more than welcome to bring along their own data to analyze during the exercises!
Sample Readings
Introductory:
Kline, Rex B. 2010. Principles and Practice of Structural Equation Modelling, 3rd Edition. Guilford Press.
Schumacker, Randall E. 2004. A Beginner’s Guide to Structural Equation Modelling. Psychology Press.
Classic:
Hayduk, Leslie. 1987. Structural Equation Modelling with LISREL—Essentials and Advances. Baltimore: Johns Hopkins University Press.
Bollen, Kenneth. 1989. Structural Equation Modelling with Latent Variables. Wiley.
Medium Difficulty:
Byrne, Barbara. 2010. Structural Equation Modelling with MPLUS: Basic Concepts, Applications, and Programming. Rutledge Academic.
More Difficult:
Kaplan, David. 2008. Structural Equation Modelling: Foundations and Extensions, 2nd edition. Sage