Distribuciones aproximadas del estadístico de Durbin-Watson para modelos de regresión no-lineal

ABRIL, JUAN CARLOS; MAEKAWA, KOICHI

Salta

Congreso; XIX Coloquio Argentino de Estadística; 1992

Sociedad Argentina de Estadística

This paper derives second and sometimes third order approximations to the density of the Durbin-Watson test statistic d for a non-linear regression model under both the null and non-null hypotheses. To derive the approximations we follow Durbin's (1980a) method for obtaining asymptotic expansions for the densities of sufficient estimators.