Selecting the Strongest Linear Relationship Between a Response Variable and an Explanatory Variable in Measurement Error Models

摘要

Measurement errors are the differences between the actual desired values and the observed values. In the real world, it is usually very difficult to obtain exactly the 'true' values. Instead, one may only get the observed values that are related to the true values through the measurement errors. In this paper we investigate the problem of selecting the treatment that has the strongest relationship between the response variable and an explanatory variable in a linear measurement error model. A selection procedure based on moment estimates has been developed and the large sample performance of the derived selection rule has also been analyzed. At the end of this paper, a simulation study is carried out to illustrate the large sample performance of the selection procedure.