摘要：Nonresponse is a pervasive and persistent problem in survey data. This research reviews several methods for imputing missing values. A special emphasis is placed on the multiple imputation methods as a more generalizable advanced remedy to missingness. An empirical application of these methods, along with a regression or conditional mean imputation, is provided. Contingent upon certain properties of data, inference based guidance to the choice and implementation of these methods is provided.
