Prepare syntax to impute the missing values using the software of your choosing. Keep in mind what type of model should be used for the imputation, depending on the type of variable to be imputed. (For example, linear regression for continuous variables, ordinal or cumulative logit regression for ordered variables, binary logistic regression for binary variables, and so forth.) If you choose to use the Stata 10 ice command for imputation, use the dryrun option first to carefully check your imputation syntax and logic. Use M = 5, or prepare five imputed data sets during this step, and be sure to use a seed value so that your results can be replicated at a later time.