Robust inference for generalized linear models
-
Cantoni, Eva
Department of Economics, University of Geneva, Switzerland
-
Ronchetti, Elvezio
Istituto di finanza (IFin), Facoltà di scienze economiche, Università della Svizzera italiana, Svizzera
Published in:
- Journal of the American Statistical Association. - American Statistical Association. - 2001, vol. 96, no. 455, p. 1022-1030
English
By starting from a natural class of robust estimators for generalized linear models based on the notion of quasi-likelihood, we de¯ne robust deviances that can be used for stepwise model selection as in the classical framework. We derive the asymptotic distribution of tests based on robust deviances and we investigate the stability of their asymptotic level under contamination. The binomial and Poisson models are treated in detail. Two applications to real data and a sensitivity analysis show that the inference obtained by means of the new techniques is more reliable than that obtained by classical estimation and testing procedures.
-
Language
-
-
Classification
-
Economics
- Other electronic version
-
Versione pubblicata
-
License
-
License undefined
-
Identifiers
-
-
RERO DOC
7820
-
ARK
ark:/12658/srd1317940
-
Persistent URL
-
https://n2t.net/ark:/12658/srd1317940
Statistics
Document views: 84
File downloads:
- ronchetti_JASA_2001.pdf: 294