I need to use the design weight for data with missing values and I am not sure when I need to weight the data.
Which way should I do:
1) weight the data.
2) run multiple imputation (MI)
1) run multiple imputation
2) weight each 5 of multiply-imputed data (5 is the default number of the subsets of MI data you get)
I will use MI data for HLM.
Thank you so much for your insights.