Date:         Wed, 24 Aug 2005 20:58:38 -0400
Reply-To:     "S. E. Hayman-Abello" <hayman2@uwindsor.ca>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         "S. E. Hayman-Abello" <hayman2@uwindsor.ca>
Subject:      multinomial logistic regression
Content-type: text/plain; charset="US-ASCII"

(second request; trying to be more succinct in the hopes of a response)

I am running a multinomial logistic regression with 2 continuous IVs , 3
categorical IVs (2 levels each), and a DV with 3 levels. I have about 120
subjects.

When I run NOMREG (using BY for the categorical variables and WITH for the
continuous variables) the output tells me that 219 cells have zero
frequencies. Also, there are 212 df listed in the goodness of fit test. I
can't make sense of these seemingly high cell numbers and hope someone can
help me.

Thanks in advance,
Sue