Imputation of Numerical Data under Edit Restrictions

02/02/2017 11:15
In this paper we propose a new imputation method satisfying edits.
Imputation is a frequently used way to treat missing data due to item-nonresponse in surveys. In some cases, especially at national statistical institutes, the imputation process is complicated due to the existence of so-called edit rules, or edits for short. Edits are constraints on the values of (combinations of) variables. Examples of edits for numerical data are that the profit of an enterprise equals its turnover minus its costs, and that the turnover of an enterprise should be at least zero.