Robust Wald-type tests for non-homogeneous observations based on minimum density power divergence estimator

摘要

This paper considers the problem of robust hypothesis testing undernon-identically distributed data. We propose Wald-type tests for both simpleand composite hypothesis for independent but non-homogeneous observations basedon the robust minimum density power divergence estimator of the commonunderlying parameter. Asymptotic and theoretical robustness properties of theproposed tests have been discussed. Application to the problem of testing thegeneral linear hypothesis in a generalized linear model with fixed-design hasbeen considered in detail with specific illustrations for its special casesunder normal and Poisson distributions.