Influence analysis for the factor analysis model with ranking data.

摘要: Influence analysis is an important component of data analysis, and the local influence approach has been widely applied to many statistical models identify influential observations assess minor model perturbations since pioneering work Cook (1986). The often adopted develop procedures for factor with ranking data. However, as this well-known based on observed likelihood, which involves multidimensional integrals, directly applying it difficult. To address difficulty, a Monte Carlo expectation maximization algorithm (MCEM) used obtain maximum-likelihood estimate parameters, measures basis conditional complete log likelihood at E-step MCEM are then obtained. Very little additional computation needed compute measures, because possible make use by-products estimation procedure. that several typical perturbation schemes discussed in detail, proposed method illustrated two real examples artificial example.