I'm running two separate LR's on my data set. I split up the analysis because two of my variables have high missingness, and I wanted to see what's going on in the greatest number of cases I can include (think 4800 rather than 1960).
When I add my variables with high missingness into a second model, the variables that are significant are much different than the ones that are significant in the original model.
It would help if I could get descriptives and frequencies for the data that is retained in the first model, and compare those to the data in the second model. Is there a way to ask SPSS to generate freq and descriptives for the cases included in a given logistic regression model?
Thanks,
Amy
------------------------------
Amy Anderson
------------------------------