Two ways of modelling overdispersion in non-normal data

For non-normal data assumed to have distributions, such as the Poisson distribution, which have an a priori dispersion parameter, there are two ways of modelling overdispersion: by a quasi-likelihood approach or with a random-effect model. The two approaches yield different variance functions for th...

Ausführliche Beschreibung

Gespeichert in:

Bibliographische Detailangaben
Veröffentlicht in:	Applied statistics 2000, Vol.49 (4), p.591-598
Hauptverfasser:	Lee, Y., Nelder, J. A.
Format:	Artikel
Sprache:	eng
Schlagworte:	Analytical estimating Binomials Data analysis Distribution Epilepsy Exact sciences and technology Generalized linear model Graph theory Hierarchical generalized linear models Linear inference, regression Linear models Mathematical models Mathematics Model checking Modeling Multilevel models Overdispersion Parametric models Probability and statistics Quasi-likelihood model Random-effect model Robust residual Sciences and techniques of general use Statistical methods Statistical models Statistical variance Statistics
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

Beschreibung
Zusammenfassung:	For non-normal data assumed to have distributions, such as the Poisson distribution, which have an a priori dispersion parameter, there are two ways of modelling overdispersion: by a quasi-likelihood approach or with a random-effect model. The two approaches yield different variance functions for the response, which may be distinguishable if adequate data are available. The epilepsy data of Thall and Vail and the fabric data of Bissell are used to exemplify the ideas.
ISSN:	0035-9254 1467-9876
DOI:	10.1111/1467-9876.00214