Maximum Likelihood Estimation of Multiple Membership Mixed Models Used in Value-Added Modeling

An EM algorithm, Karl et al. (2013) , estimates generalized, reduced generalized, variable, complete, and zero persistence models by maximum likelihood or restricted maximum likelihood (the default). These are multiple-membership linear mixed models with teachers modeled as "G-side" effects and students modeled with either "G-side" or "R-side" effects; see Mariano et al. (2010) .