When there are outliers in the carriers:the univariate case
- 1 January 1982
- journal article
- research article
- Published by Taylor & Francis in Communications in Statistics - Theory and Methods
- Vol. 11 (8) , 849-868
- https://doi.org/10.1080/03610928208828277
Abstract
Robust regression estimators studied to date are robust against non-normal distributions of the errors only If the carriers ‘Independent variables’ do not also contain outliers. Several alternative estimators that are robust even 1f there are outliers in the carriers are studied. Two estimators seem to be preferable, but even these can be very Inefficient ‘relative to least squares’ If the errors are normally distributed.Keywords
This publication has 10 references indexed in Scilit:
- A Comparative Study of Several Robust Estimates of Slope, Intercept, and Scale in Linear RegressionJournal of the American Statistical Association, 1977
- Confidence Intervals for Bisquare Regression EstimatesJournal of the American Statistical Association, 1977
- A Simple Method for Robust RegressionJournal of the American Statistical Association, 1975
- Adaptive Robust Procedures: A Partial Review and Some Suggestions for Future Applications and TheoryJournal of the American Statistical Association, 1974
- A Robust Method for Multiple Linear RegressionTechnometrics, 1974
- The Influence Curve and its Role in Robust EstimationJournal of the American Statistical Association, 1974
- The Fitting of Power Series, Meaning Polynomials, Illustrated on Band-Spectroscopic DataTechnometrics, 1974
- Robust Regression: Asymptotics, Conjectures and Monte CarloThe Annals of Statistics, 1973
- Monte Carlo for Robust Regression: The Swindle UnmaskedPublished by National Bureau of Economic Research ,1973
- A General Qualitative Definition of RobustnessThe Annals of Mathematical Statistics, 1971