期刊名称：Brazilian Journal of Probability and Statistics
印刷版ISSN：0103-0752
出版年度：2006
卷号：20
期号：2
页码：179-190
出版社：Brazilian Statistical Association
摘要：In this paper, we introduce the multivariate skew-normalmodel in the context of measurement error mo dels in order to avoid datatransformations or usual constraints on the parametric space. This distribu-tion was recently discussed by Capitanio, Azzalini and Stanghellini (2003)using graphical models. The motivation is to use the conditioning argumenton the unobserved true value of the explanatory variable in the two-variablemeasurement error model in order to get the skewness parameters as a func-tion the original parameters of the proposed measurement error models. In-ferential problems are considered from the Bayesian point of view assumingproper noninformative priors via Winbugs. The usefulness of the proposedmodel with errors in variables is investigated with a simulation study and realdata analysis. The main advantage of the Bayesian approach is the possibilityto measure the degree of belief that the true value of the explanatory variableis greater than its mean value. This constraint implies a strong asymmetryon the distribution of the unobserved true value. We end the paper by con-cluding that the extended skew-normal measurement error model provides.exibility in terms of skewness without making any additional assumptionsto eliminate the usual identifiable problems in the measurement error models
