scientific article; zbMATH DE number 19006

MaRDI QIDQ3976431zbMATH OpenFDO

Authors Marianthi Markatou, Werner A. Stahel, Elvezio Ronchetti

Publication date 26 June 1992

zbMATH Keywords

\(M\)- estimatorsaffine equivariance propertyasymptotic powerbreakdown analysisdrop-in-dispersion testsinfluence functionlinear modelMonte Carlo resultsnonrobust estimatorsoptimal robust testsrobust testing proceduresscores testsmall sample behaviourstandardization matricesWald test

Mathematics Subject Classification ID

Asymptotic properties of parametric tests (62F05) Robustness and adaptive procedures (parametric inference) (62F35) Linear regression; mixed models (62J05)

Review text

[For the entire collection see Zbl 0731.00018.] After a short overview of robust testing procedures for the linear model that are based on \(M\)-estimators and related statistics, three classes of tests, the Wald, scores, and drop-in-dispersion tests are introduced and their asymptotic properties are investigated. Optimal robust tests in the sense of maximizing the asymptotic power within a class and subject to a bound of the influence function are given. In order to obtain tests as well as estimators with the usual affine equivariance property standardization matrices are needed. The authors give a general idea how some nonrobust estimators of such matrices can be robustified. It relies on downweighting observations with large residuals from a preliminary robust fit. Finally, some Monte Carlo results on the small sample behaviour of the three classes of the tests are reported from the literature and some remarks on breakdown analysis are made.

Recommendations

• Robust Testing Procedures in Heteroscedastic Linear Models
• Bounded Influence and High Breakdown Point Testing Procedures in Linear Models
• Robust Bounded-Influence Tests in Linear Models
• Robust Wald-Type Tests of One-Sided Hypotheses in the Linear Model

Cited in

(16)

• Fast and robust bootstrap
• Robust inference by influence functions
• Robust estimators for simultaneous equations models
• On the diversity of estimates.
• Robust tests for linear regression models based on \(\tau\)-estimates
• Estimating the \(p\)-values of robust tests for the linear model
• Optimum robust testing in linear models
• On the robustness of mixture index of fit
• Influence analysis of robust Wald-type tests
• Robust Bounded-Influence Tests in Linear Models
• scientific article; zbMATH DE number 6511319 (Why is no real title available?)
• Robust Testing Procedures in Heteroscedastic Linear Models
• scientific article; zbMATH DE number 4066140 (Why is no real title available?)
• scientific article; zbMATH DE number 926557 (Why is no real title available?)
• Simple powerful robust tests based on sign depth
• M-tests in multivariate models

This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3976431)