التعـويض الجزئي للقيم المفقودة

Abstract

If the problem of missing values, during the data collection stage, happened completely at random (MCAR), we can simply use the deletion method to solve it. But if the missing happened at random (MAR), we can often make it as MCAR by partial imputation. That is based on replacing estimated values with some of the missing values instead of all of them, using one of the single imputation methods for simplicity, which are available in most of the statistical packages. Then we can use the deletion method as a solution. The simplicity of these steps enables unspecialized researcher to imply them alone, without asking help, which is usually hard to obtain.