A note on efficient regression estimators with positive breakdown point

In this note we prove that for certain types of regression estimators with positive breakdown point, the finite sample efficiencies can be arbitrarily low when the design is unfavorable. In particular, the estimator proposed by Yohai and Zamar (1988), which is asymptotically efficient when the errors are normal, falls in this category.

MoreLess

Year of publication:	1991
Authors:	Morgenthaler, S.
Published in:	Statistics & Probability Letters. - Elsevier, ISSN 0167-7152. - Vol. 11.1991, 6, p. 469-472
Publisher:	Elsevier

More details

Type of publication:	Article
Source:	RePEc - Research Papers in Economics

Persistent link: https://www.econbiz.de/10005319347