Incomplete Data Filled

In the field of each research and investigation, because of many known or unknown factors, we often have to face missing data. Then, because of the missing data, it not only let the analysis became more and more difficult but also cause the error of results, then decrease the effect of statistical work. So how to think of the efficiency of these missing data become very important. In order to solve the problem, we introduce a method, that is: impute the missing data and give the complete data sets, then analyze them, get the final statistical results.This paper just introduce the theory of the imputation of missing data, from simple imputation to multiple imputation, we introduce mean imputation, regression imputation, multiple imputation, Markov chain Monte Carlo(MCMC) in detail and analyze their difference. At last, we use SAS software to explain the different methods with an example. In the meanwhile, we introduce the concept of group decision, use the weighted method on group decision to solve the example.