Page 1 of 1
R2MLwiN - cross-classified with IGLS?
Posted: Mon Nov 18, 2013 1:19 am
by danielledean527
I'm having trouble figuring out how to run both cross-classified and multiple membership models using estimation methods other than MCMC with R2MLwiN - are these models only estimable with the R package with MCMC? If not, can anyone provide an example? Thanks!
Re: R2MLwiN - cross-classified with IGLS?
Posted: Tue Nov 19, 2013 2:40 pm
by richardparker
Hi - yes, at the moment it's only possible to fit cross-classified and multiple membership models via MCMC estimation when using R2MLwiN.
For more information on the features R2MLwiN supports for IGLS and MCMC estimation, see ?MacroScript1 (the R2MLwiN function which writes out the MLwiN macro for IGLS estimation) and ?MacroScript2 (the R2MLwiN function which writes out the MLwiN macro for MCMC estimation).
Incidentally, if you want to fit cross-classified models, in R, using non-Bayesian methods, then a package such as lme4 would be a good choice.