The SURVEYIMPUTE procedure provides the following donor-based imputation methods for handling missing values in survey data: single and multiple hot-deck imputation, approximate Bayesian bootstrap imputation, fully efficient fractional imputation, two-stage fully efficient fractional imputation, and fractional hot-deck imputation. When you use fully efficient fractional imputation or fractional hot-deck imputation, PROC SURVEYIMPUTE produces replicate weights that appropriately account for the imputation. You can use these replicate weights in any survey analysis procedure to correctly estimate the variance for an estimator that uses the imputed data. For more information about fully efficient fractional imputation, see Kim and Shao (2014, sections 8.4 and 8.5) and Fuller (2009, section 5.2).
For more information, see Chapter 119, The SURVEYIMPUTE Procedure.