Meng and Schilling illustrate two ways of implementing the Monte Carlo Expectation Maximization algorithm to fit a full-information item factor model, using the Gibbs sampler to carry out the computation for the “E” steps.