Hi all,
I got confused when I tried to run the data containing some missing values in the response.
When I run it with the missing values, it is said that "14394 of 14472 cases in use", what I undersatnd from it is that the program only make use of the non-missing data to build the model, and I suppose the result would be the same if I omit the missing cases and run the same model again, but it is not the case. When I got the 14394 non-missing cases only, the estimates of the coefficients are different!!
Is there anyone know what does "14394 of 14472 cases in use" really mean if it is not as what I understood?
Thank you all.
With best regards,
Pida