On local influence analysis of full information item factor models

摘要: The full information item factor (FIIF) model is very useful for analyzing relations of dichotomous variables. In this article, we present a feasible procedure to assess local influence minor perturbations identifying aspects the FIIF model. development based on Q-displacement function which closely related with Monte Carlo EM algorithm in ML estimation. E-step algorithm, conditional expectations are approximated by sample means observations simulated Gibbs sampler from appropriate distributions. It turns out that these can be utilized computing building blocks proposed diagnostic measures. diagnoses conformal normal curvature computed easily. A number interesting perturbation schemes considered. methodology illustrated two real examples.