Generalized Linear Latent and Mixed Modelling (GLLAMM): The framework, estimation and some examples
Pre-conference workshop with Andrew Pickles, University of Manchester, England
20th July, 14.30- 18.30, Castlereagh Room, Fisher Building
I introduce the GLLAMM framework, which subsumes an enormous range of multivariate random effect, latent variable and latent class models for response variables from the generalized linear model family, and best set out in Skrondal & Rabe-Hesketh (2004). I describe model estimation and post-estimation using gllamm, the procedure implemented in Stata (www.gllamm.org). Some examples will be chosen to illustrate a range of features. These will include multilevel IRT modelling, the use of trajectory models in intervention/causal effect estimation studies, and expanded or composite link models.
References
Skrondal, A. and Rabe-Hesketh, S. (2004). Generalized Latent Variable Modeling: Multilevel, Longitudinal and Structural Equation Models. Chapman & Hall/CRC
Rabe-Hesketh, S., Skrondal, A. and Pickles, A. (2004). GLLAMM Manual U.C. Berkeley Division of Biostatistics Working Paper Series. Working Paper 160.