Performance Comparison of Imputation Methods for Mixed Data Missing at Random with Small and Large Sample Data Set with Different Variability

One of the concerns in field statistics is presence missing data, which leads to bias parameter estimation and inaccurate results. However, multiple imputation procedure a remedy for handling data. This study looked at best methods used handle mixed variable datasets with different sample sizes variability along levels missingness. The employed predictive mean matching, classification regression trees, random forest methods. For each dataset, estimates complete were compared found imputed dataset. results showed that method was mostly 500 irrespective variability. tree worked on 30