Dear statalist,

I'm running a logistic regression and there are two independent variables with missing values. So I first include only those subjects without missing values and second include all subjects after imputation. If I want to do sensitivity analysis for this two results, which method should I use? Thank you!