Artículos de revistas
Model for distributions and correlation structure of multivariate polytomous data from a two-stage cluster sample
Registro en:
Communications In Statistics-simulation And Computation. Marcel Dekker Inc, v. 25, n. 1, n. 25, n. 39, 1996.
0361-0918
WOS:A1996TY60900003
10.1080/03610919608813296
Autor
Marques, EHD
Koch, GG
Institución
Resumen
A generalization of the methods for analyzing bivariate dichotomous data from a stratified two-stage cluster sample (Marques and Koch, 1992) is presented in this paper. This development includes the joint estimation of means for marginal distributions and a set of correlation measures for intra-cluster association for the respective outcomes of each response criterion and for correlation among response criteria for units sharing the same cluster in a multivariate polytomous data setting. The methodology is illustrated with a medical example of clinical effects of topical corticosteroids on healthy skin. 25 1 25 39