An assessment of the pseudo-likelihood estimators
for the spatial autologistic model

Denise Nunes VIOLA[1]

Clarice Garcia Borges DEMÉTRIO[2]

Paulo Justiniano RIBEIRO JR[3]

Bryan Frederick John MANLY[4]

§    ABSTRACT: In this paper a simulation study on pseudo-likelihood estimators of autologistic parameters to verify the effect of different covariate and neighbouring structures is described, with three disease levels and five different spatial correlation coefficient values. An application of the methodology is presented using bell pepper data from Gumpertz, Graham and Ristaino~(1997). It is shown that the pseudo-likelihood method can be used when a researcher is interested in the effect of covariates, but should not be used for the estimation of the spatial correlation.

§    KEYWORD: Autologistic model; spatial correlation; binary data; pseudo-likelihood; bootstrap.

 



[1] Departamento de Estatística, Universidade Federal da Bahia, Av. Adhemar de Barros, s/n - Campus de Ondina CEP: 40.170-110, Salvador, BA, Brasil. E-mail: viola@ufba.br

[2] Departamento de Ciências Exatas, ESALQ/USP, Avenida Pádua Dias, 11, CEP: 13418-900, Piracicaba, SP, Brasil. E-mail: clarice@esalq.usp.br

[3] Laboratório de Estatística e Geoinformação, Universidade Federal do Paraná, Caixa Postal 19.081, CEP: 81531-990, Curitiba, PR, Brasil. E-mail: paulojus@ufpr.br

[4] Departamento de Ciências Exatas, ESALQ/USP, Avenida Pádua Dias, 11, CEP: 13418-900, Piracicaba, SP, Brasil. E-mail: bryanmanly@lycos.com