Biases and Bartlett and Bartlett-type corrections in extreme value linear regression models

Edson Marcos Leal Soares RAMOS[1]

Gauss Moutinho CORDEIRO[2]

§     ABSTRACT: In this paper, we obtain bias corrected maximum likelihood estimates in extreme value linear regression models. We also consider finite-sample corrections in matrix notation for the likelihood ratio and score statistics in these models. Some simulations are performed to compare the likelihood ratio and score statistics and their modified versions with the chi-squared distribution. We also illustrate the bias correction.

§     KEYWORDS: Bias correction; bootstrap test; chi-squared distribution; extreme value model; maximum likelihood estimate.

 



[1]Departamento de Estatística, Universidade Federal do Pará - UFPA, Caixa postal 479, CEP 66075-110, Belém, PA, Brasil. E-mail: ramosedson@yahoo.com.br

[2] Departamento de Estatística e Informática, Universidade Federal Rural de Pernambuco - UFRPE, CEP 50171-900, Recife, PE, Brasil. E-mail: gauss@ufrpe.br