www.cmm.bristol.ac.uk/forum

I am trying to impute missing values in a multilevel model. The program seems to run appropriately and outputs the specified imputed datasets. At this point, I need to import the datasets into SAS. Since the imputed datasets only include the response variables and are missing subject-level identifiers, I am not sure how to then merge the imputed dataset back with the original data. Is there a way to either have the imputed datasets include all of the variables included in the model (both auxiliary and response) or include a subject number in the imputed dataset?

Thanks for your help!

There are currently no options for specifying how the imputed data is saved from Realcom, however as the data should be in the same order as your original file you could generate a variable based on the record number in both files and then merge on this.