Jackknifing Poisson Log-Linear Models

Management of Organizations: Systematic Research (Organizaciju Vadyba: Sisteminiai Tyrimai), issue 36, pp. 33-42, 2005

Posted: 3 Apr 2009

See all articles by Martin Knott

Martin Knott

affiliation not provided to SSRN

Christos C. Frangos

Technological Educational Institute of Athens - Department of Business Administration

Date Written: 2005

Abstract

We investigate the performance of jackknife estimators based on maximum likelihood estimators of log-liner models for Poisson observations. We find that though for very simple models jackknifing may improve both bias and variance, in general there is no such strong optimum. We suggest new more accurate estimates for the variances of the estimators.

Keywords: Logit, Variance estimation, Influence function, Jackknife estimator, Log-linear Model

JEL Classification: C00

Suggested Citation

Knott, Martin and Frangos, Christos C., Jackknifing Poisson Log-Linear Models (2005). Management of Organizations: Systematic Research (Organizaciju Vadyba: Sisteminiai Tyrimai), issue 36, pp. 33-42, 2005, Available at SSRN: https://ssrn.com/abstract=1372614

Martin Knott

affiliation not provided to SSRN

Christos C. Frangos (Contact Author)

Technological Educational Institute of Athens - Department of Business Administration ( email )

Agiou Spyridonos Street
Aigaleo
Athens, Attiki 12210
Greece
+302105385214 (Phone)

HOME PAGE: http://users.teiath.gr/cfragos

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Abstract Views
586
PlumX Metrics