Dealing with non-ignorable sampling and non-response in statistical matching

Statistical matching deals with the problem of how to combine information collected in different samples taken from the same target population, but on partly different survey variables. In this article, we consider the use of statistical matching when the samples are drawn by informative sampling designs and are subject to not missing at random non-response. The problem with ignoring the sampling process and non-response can result in severely biased estimators and a misrepresentative fused dataset. The proposed methodology employs the empirical likelihood approach.