Multivariate linear ultrastructural relationships model

Abstract

Multivariate errors-in-variables model in which the unobservable true values are assumed to be random variables satisfying linear relationships is considered. When these unobservables are normally distributed but with distinct location parameters, it is called the multivariate linear ultrastructural relationships model. Maximum likelihood estimation of the parameters of this model is considered here assuming repeated observations.

Keywords:

• Errors-in-variables
• functional relationships
• structural relationships
• maximum likelihood estimation