Record number :
1557805
Title of article :
Measurement Error Models with Nonconstant Covariance Matrices
Author/Authors :
Arellano-Valle، نويسنده , , Reinaldo B. and Bolfarine، نويسنده , , Heleno and Gasco، نويسنده , , Loreta، نويسنده ,
Issue Information :
دوفصلنامه با شماره پیاپی سال 2002
Pages :
21
From page :
395
To page :
415
Abstract :
In this paper we consider measurement error models when the observed random vectors are independent and have mean vector and covariance matrix changing with each observation. The asymptotic behavior of the sample mean vector and the sample covariance matrix are studied for such models. Using the derived results, we study the case of the elliptical multiplicative error-in-variables models, providing formal justification for the asymptotic distribution of consistent slope parameter estimators. The model considered extends a normal model previously considered in the literature. Asymptotic relative efficiencies comparing several estimators are also reported.
Keywords :
Asymptotic distribution , Elliptical distribution , multiplicative measurement error model , sample mean vector and sample covariance matrix
Journal title :
Journal of Multivariate Analysis
Serial Year :
2002
Link To Document :
بازگشت