$4.$ The reason why multiple imputation may be preferred to single imputation is: a$.$ It makes the data scientist look good. b$.$ It results in the proper level of variability in the imputed value to support statistical inference. c$.$ It addresses the issue of missing completely at random while simple imputation does not. d$.$ It addresses the issue of missing and extreme values while simple imputation only addresses missing values. e$.$ All the above are true.