On Pearson's residuals in generalized linear models
A rigorous asymptotic theory for Pearson residuals in generalized linear models is not yet available. We give matrix formulae of order n-1, where n is the sample size, for the first two moments of these residuals. The formulae are applicable to many regression models in common use. We suggest adjusted Pearson residuals in these models with approximately zero mean and unit variance.
Year of publication: |
2004
|
---|---|
Authors: | Cordeiro, Gauss M. |
Published in: |
Statistics & Probability Letters. - Elsevier, ISSN 0167-7152. - Vol. 66.2004, 3, p. 213-219
|
Publisher: |
Elsevier |
Keywords: | Exponential family Generalized linear model Link function Log-linear model Pearson residual |
Saved in:
Saved in favorites
Similar items by person
-
Corrected maximum likelihood estimators in linear heteroskedastic regression models
Cordeiro, Gauss M., (2008)
-
Corrected maximum likelihood estimators in linear heteroskedastic regression models
Cordeiro, Gauss M., (2008)
-
Discussion of “Birnbaum‐Saunders distribution : A review of models, analysis, and applications”
Lemonte, Artur J., (2018)
- More ...