Events and Meetings of Italian Statistical Society, Advances in Latent Variables - Methods, Models and Applications

Font Size: 
A weighted pairwise likelihood estimator for a class of latent variable models
Vassilis Vasdekis, Dimitris Rizopoulos, Irini Moustaki

Last modified: 2013-06-16

Abstract


Random effects and latent variable models are widely used for capturing unobserved heterogeneity in longitudinal studies and associations in multivariate data. The estimation of those models becomes cumbersome as the number of the random effects or latent variables increase due to the high-dimensional integrations involved. Composite likelihood is a pseudo-likelihood that combines lower-order marginal or conditional densities such as univariate or bivariate and it has been proposed in the literature as an alternative to full maximum likelihood estimation. We propose a weighted pairwise likelihood estimator based on estimates obtained from separate maximizations of marginal pairwise likelihoods. The derived weights minimize the total variance of the estimated parameters. The proposed weighted estimator is found to be more efficient than the one that assumes all weights to be equal.

Full Text: PDF