SFB 303 Discussion Paper No. B-326
Author: Werner, Hans Joachim, and Cemil Yapar
Title: A BLUE Decomposition in the General Linear Regression Model
Abstract: In this note we consider the {\it general} linear regression model
$(y,\ X\beta,\ V\mid\ R_2\beta_2=r)$ where the block partitioned
regressor matrix $X=\pmatrix{X_1 & X_2\cr}$ may be deficient in
column rank, the dispersion matrix $V$ is possibly singular,
$\beta^t=\pmatrix{\beta_1^t & \beta_2^t\cr}$ - being partitioned
according to $X$ - is the vector of unknown regression coefficients,
and $\beta_2$ is possibly subject to consistent linear constraints
$R_2\beta_2=r$. Of much interest to us is the {\it traditional} BLUE
{\it (best linear unbiased estimator)} of $X\beta$. We show how this
traditional BLUE can obtained from the traditional BLUE of
$X_1\beta_1$ in the related model $(y,\ X_1\beta_1,\ V)$. Some
properties of the dispersion matrix of the traditional BLUE of
$X\beta$ are also given.
Keywords: Linear regression model, rank deficiency, linear
constraints, traditional BLUE, dispersion, g-inversion.
JEL-Classification-Number: C20
Creation-Date: 1995
Unfortunately this paper is not available. Please order a hardcopy vial
mail.
17.02.1998, © Webmaster
