Premium
INFLUENCE DIAGNOSTICS IN PROPER DISPERSION MODELS
Author(s) -
Paula Gilberto A.
Publication year - 1996
Publication title -
australian journal of statistics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.434
H-Index - 41
eISSN - 1467-842X
pISSN - 0004-9581
DOI - 10.1111/j.1467-842x.1996.tb00685.x
Subject(s) - univariate , dispersion (optics) , econometrics , computer science , displacement (psychology) , generalized linear model , regression analysis , linear regression , statistics , mathematics , multivariate statistics , physics , optics , psychology , psychotherapist
Summary This paper discusses the application of influence diagnostic methods in univariate proper dispersion models. This class includes, in particular, continuous generalized linear models as well as other subclasses of continuous regression models. We emphasize the study of the local influence on the likelihood displacement and predictions from the models. Some of the diagnostics are illustrated by an example on directional data.