Prev: Help in PROC compare needed.
Next: Recoding All variables in a dataset
From: Tristan Berckmans on 6 May 2010 09:13
I have done logistic regression on my data: the catmod procedure.
Using accounting numbers as independent variables, I am trying to
estimate if a firm either purchases assets, does nothing or sells
assets.
The likelihood ratio test can't handle my continous independent
variables and groups the 484 observations as different populations. I
wonder if there is another statistic to give the goodness-of-fit of my
model, a kind of pseudo R²? One that does not care about my continuous
dependent variabeles?

I appreciate your time and help!
 | 
Pages: 1
Prev: Help in PROC compare needed.
Next: Recoding All variables in a dataset