Statistical Modeling in Mplus

March 12 - 16, 2012  

Lecturer:

Prof. Dr. Joop Hox, Utrecht University, NL  

Abstract:

Structural equation modeling (SEM) is a very general technique for multivariate analysis, which includes techniques as multiple regression and factor analysis as special cases. Originally, structural equation modeling was restricted to modeling the covariance structure of continuous variables using latent variables and path models. Over the years, structural equation modeling has been expanded to include categorical and other non-normal variables, models with means and intercepts, multilevel models, and models with categorical latent variables (latent class models). Moreover, several estimation techniques have been developed to accommodate incomplete data.   The orientation of the course is an applied one, with lectures in the morning and lab sessions in the afternoon.   The lab sessions will use the software Mplus, which is a very general package for advanced structural equation modeling. The course starts with an introduction of structural equation modeling with some emphasis on the specific way Mplus is used to specify and estimate models.  

Requirements:

The course requires a basic understanding of structural equation modeling, on the level of e.g. Kline (2011). The basics of structural equation modeling will be briefly reviewed, but not in the form of a complete introduction.  

Basic Reading:

General SEM:

• Kline, R. (2011). Principles and Practice of Structural Equation Modeling. London.
• Hancock, G. R. & R. O. Mueller (2006). Structural Equation Modeling: A Second Course. Greenwich.  

Specific Mplus:

• Byrne, B.M. (2012). Structural Equation Modeling with Mplus. New York. §  Geiser, C. (2010). Datenanalyse mit Mplus. Wiesbaden.