I wish to conduct sensitivity analysis on some outcome variables which have missing data.
I want to try to assign my patients with missing data to one of the binary outcomes (e.g. yes or no), then repeat univariate analysis to see any changes. I believe this is a way to see if my data is missing at random or not.
I have been unsuccessful at identifying this type of method so I haven't been able to test it out!
Does anyone have any advice or code that could help me please?