Modelos factoriales para datos longitudinales con estructuras comunes y especificas

Metodología de las Ciencias del Comportamiento 01/2004; Sup..


One of the most useful techniques to analyse latent structures is exploratory Factor Analysis. When several data matrices, arising from different occasions and/or groups, are obtained, the comparison and integration of the resulting latent spaces is an important task. In this paper we propose a factor model for several groups that includes shared common factors for all the groups and specific factors for each group. Shared common factors are obtained from the common directions of the separated latent subspaces and specific factors from the residuals after projecting the group onto the common subspace

This research doesn't cite any other publications.