Grouping and Rank Estimators in EVMS

This paper first discusses various grouping/rank estimation methods in the EVMs. In doing so it gives an exposition of some of the recent work done in this area and compares them. An improved estimator is proposed and is compared with other estimators. It is shown that this estimator is uniformly be...

Full description

Saved in:
Bibliographic Details
Published in:Sankhyā. Series B 1984-04, Vol.46 (1), p.90-107
Main Authors: Bhaumik, Mrinal, Pal, Manoranjan
Format: Article
Language:English
Subjects:
Consistent estimators
Estimation bias
Estimation methods
Estimators
Instrumental variables estimation
Least squares
Mathematical independent variables
Regression analysis
Regression coefficients
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper first discusses various grouping/rank estimation methods in the EVMs. In doing so it gives an exposition of some of the recent work done in this area and compares them. An improved estimator is proposed and is compared with other estimators. It is shown that this estimator is uniformly better than the other estimators in the sense of minimizing asymptotic variance within the class of estimators. Second, to prove an important result of the paper, it is assumed that the ranks of the true values of the regressor are known from an independent source. In this case, an optimum minimum variance rank estimator is found. The optimisation of the class of estimators proposed in this paper leads to a simple and elegant approach which can easily be applied in other similar situations.
ISSN:0581-5738