Structural Equation Modelling in Mplus

Dates: 22nd March 2010
Duration: 1 day
Level: Intermediate
Course Fee: £175 (£125 for those from educational institutions)

CCSR offer 5 free places to research staff and students within the Faculty of Humanities at the University of Manchester.
Course Leader: Nick Shryane

Course Requirements: Participants should be familiar with regression models for continuous and binary data. No experience with the Mplus package is assumed.

 

Summary

This one-day course provides an introduction to basic concepts and models in structural equation modeling (SEM) and ab initio training in using the Mplus statistical software package to fit these models. The course is intended as a practical introduction focusing on the conceptual basis of SEM and social science applications. 

 

Objectives

The completion participants should have understanding of:

• Latent variables / latent trait / factor models
• Measurement vs. Structural models
• How to evaluate and compare model fit

 

And be able to use Mplus to specify and evaluate the following models for cross-sectional data:

• Confirmatory Factor Analysis models
• Item Response Theory models
• Covariate measurement error models
• Mediation models

 

Target Audience

Social Scientists who are familiar with commonly used regression techniques but who want address complex social research questions in a single framework and address the issues of construct measurement and measurement error.

 

Preliminary Reading

If you wish to prepare for the course then there is a chapter on SEM in this popular general textbook:

Tabachnick & Fidell (2007). Using Multivariate Analysis Statistics, 5th Edition. Pearson Education.

This is a dedicated book covering many aspects of SEMs:

Loehlin (2004). Latent Variable Models, 4th Edition. Erlbaum.