Performances d'estimateurs à rétrécisseur en situation de multicolinéarité

Christian Robert

Annals of Economics and Statistics, 1988, issue 10, 97-119

Abstract: Near collinearities among explicative variables in a regression model have unwanted effects on the least squares estimator. They inflate the variances of least squares regression coefficient estimates and introduce a lack of fiability for this estimator. In this paper, we define precisely the notion of multicollinearity, then we show why a minimax shrinkage estimator cannot bring any significant improvement over the LSE in the estimation of the components responsible for multicollinearity. In the last part, we propose a generalization of the principal components estimators which performs rather well in a bounded neighborhood of 0.

Date: 1988
References: Add references at CitEc
Citations:

Downloads: (external link)
http://www.jstor.org/stable/20075697 (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:adr:anecst:y:1988:i:10:p:97-119

Access Statistics for this article

Annals of Economics and Statistics is currently edited by Laurent Linnemer

More articles in Annals of Economics and Statistics from GENES Contact information at EDIRC.
Bibliographic data for series maintained by Secretariat General () and Laurent Linnemer ().