Robust Tests in Generalized Linear Models with Missing Responses

Computational Statistics & Data Analysis, 65 (2013), 80-97
http://dx.doi.org/10.1016/j.csda.2012.05.008

In many situations, data follow a generalized linear model in which the mean of the responses is modelled, through a link function, linearly on the covariates. Robust estimators for the regression parameter in order to build test statistics for this parameter, when missing data occur in the responses, are considered. The asymptotic behaviour of the robust estimators for the regression parameter is obtained, under the null hypothesis and under contiguous alternatives. This allows us to derive the asymptotic distribution of the robust Wald-type test statistics constructed from the proposed estimators. The influence function of the test statistics is also studied. A simulation study allows us to compare the behaviour of the classical and robust tests, under different contamination schemes. Applications to real data sets enable to investigate the sensitivity of the pp-value to the missing scheme and to the presence of outliers.