Facoltà di scienze economiche
Texte intégral

Variable selection for marginal longitudinal generalized linear models

Cantoni, Eva ; Mills-Flemming, Joanna ; Ronchetti, Elvezio

In: Biometrics, 2005, vol. 61, no. 2, p. 507-514

Variable selection is an essential part of any statistical analysis and yet has been somewhat neglected in the context of longitudinal data analysis. In this paper we propose a generalized version of Mallows's Cp (GCp) suitable for use with both parametric and nonparametric models. GCp provides an estimate of a measure of model's adequacy for prediction. We examine its performance with popular... Plus

Ajouter à la liste personnelle
    Titre
    • Variable selection for marginal longitudinal generalized linear models
    Auteur
    • Cantoni, Eva. Department of Econometrics, University of Geneva, Switzerland
    • Mills-Flemming, Joanna. Department of Econometrics, University of Geneva, Switzerland
    • Ronchetti, Elvezio. Istituto di finanza (IFin), Facoltà di scienze economiche, Università della Svizzera italiana, Svizzera
    Type de document
    • Postprint
    Date de publication
    • 2003
    Identifiant OAI-PMH
    • oai:doc.rero.ch:20070413163737-ZQ
    Numéro de soumission
    • 20070413163737-ZQ
    Summary
    Variable selection is an essential part of any statistical analysis and yet has been somewhat neglected in the context of longitudinal data analysis. In this paper we propose a generalized version of Mallows's Cp (GCp) suitable for use with both parametric and nonparametric models. GCp provides an estimate of a measure of model's adequacy for prediction. We examine its performance with popular marginal longitudinal models (fitted using GEE) and contrast results with what is typically done in practice: variable selection based on Wald-type or score-type tests. An application to real data further demonstrates the merits of our approach while at the same time emphasizing some important robust features inherent to GCp.