ABSTRACT
P. De Boeck and M. Wilson have described a really intuitive formulation of the Rasch model as a linear mixed model. Regarding the Marginal Maximum Likelihood estimation, it is typically assumed that the random effects are each normally distributed. A major characteristic of the formulation just given is the distinction between random and fixed effects. One of the most practical papers regarding the use of lme4 for the application of Rasch models in the form of generalized linear mixed model (GLMM) is that of De Boeck et al. The chapter draws on the examples of that paper, to run a Rasch model, in the form of GLMM, using explanatory variables. This way, the model resembles a typical regression model.
