Maximum Likelihood and Semiparametric Estimation in Logistic Models with Incomplete Covariate Data.
Collaborative Research Center 386, Discussion Paper 110
Maximum likelihood estimation of regression parameters with incomplete covariate information usually requires a distributional assumption about the concerned covariates which implies a source of misspecification. Semiparametric procedures avoid such assumptions at the expense of efficiency. A simulation study is carried out to get an idea of the performance of the maximum likelihood estimator under misspecification and to compare the semiparametric procedures with the maximum likelihood estimator when the latter is based on a correct assumption.