Multivariate Linear Regression Model with Elliptically Contoured Distributed Errors and Monotone Missing Dependent Variables

Abstract

In this article, the multivariate linear regression model is studied under the assumptions that the error term of this model is described by the elliptically contoured distribution and the observations on the response variables are of a monotone missing pattern. It is primarily concerned with estimation of the model parameters, as well as with the development of the likelihood ratio test in order to examine the existence of linear constraints on the regression coefficients. An illustrative example is presented for the explanation of the results.

Keywords:

• Consistency of estimators
• Elliptically contoured distributions
• Estimation
• Generalized Wilk's distribution
• Hypothesis testing
• Monotone missing data
• Multivariate linear regression analysis

Mathematics Subject Classification:

• Primary 62J12
• Secondary 62H12, 62H15

Acknowledgments

This research was partially co-funded by the European Union-European Social Fund and National Resources-EPEAEK II.