An Introduction to Latent Variable Modelling

About

This short-course provides an introduction to latent variable modelling – an ever increasingly used approach in the behavioural and social sciences.

Course leader

Professor Mark Shevlin

Target group

It is expected that participants will have some knowledge of different variable types (nominal, ordinal, etc.), descriptive statistics and a working knowledge of hypothesis testing prior to taking the course. An understanding of regression and correlation would be a benefit.

Course aim

This short-course provides an introduction to latent variable modelling – an ever increasingly used approach in the behavioural and social sciences. The short-course covers many of the major features of latent variable models including confirmatory factor analysis, path analysis (with and without error) and modelling the relationships between latent variables. The historical and statistical foundations of latent variable models will be detailed, with particular attention paid to the issues of measurement, specification, estimation and interpretation of models. We will demonstrate how latent variable models offer an extremely flexible framework for statistical analysis and one that allows complex hypotheses to be tested.

Some extensions to the basic latent variable model will be introduced, such as multiple group analysis, MIMIC models, and the application of model constraints. The course will be delivered by means of lectures and hands-on practical work. The final session of each day will include a question and answer session with the opportunity for everyone to discuss their research interests and their own data structures. Mplus will be used, but no experience of this software is required.

Fee info