Lets assume there are 150 variables in a Dataset. For my analysis I only use around 50 of them. Of those 50 variables, there are 18 Variables which contain missing values. How should I perform the Little´s test? Should I include all 150 variables or the 50 I want to use later on?
I even saw some examples where people gave only a subset of similiar variables into the test. Is that a correct way to do that?
