Inference in high dimensional linear measurement error models (Q2034474)

This article develops statistical inference on the parameters associated with the error-prone covariates in the high dimensional linear model with a finite number of covariates measured with errors. A a new corrected decorrelated score test and a corresponding score type estimator are proposed. It is shown that the limiting distribution of the new corrected decorrelated score test statistic has a standard normal distribution under the null hypothesis and retains power under the local alternatives around zero. Asymptotic normality of the new proposed estimator is also proved, so that asymptotic confidence intervals can be constructed. The finite-sample performance of the proposed inference procedure is examined through simulation studies. An application to real data collected in a clinical trial designed to determine the long-term effects of different inhaled treatments for mild to moderate childhood asthma, where phenotypic information and genome-wide SNP data are accessible, is presented.