Date:         Sat, 21 Dec 2002 21:55:02 +1000
Reply-To:     Bob Green <bgreen@dyson.brisnet.org.au>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Bob Green <bgreen@dyson.brisnet.org.au>
Subject:      Re: Testing individual regressors in a LogReg analysis
In-Reply-To:  <se0335e3.096@office.uncg.edu>
Content-Type: text/plain; charset="us-ascii"; format=flowed

Mark,

Version 10 contains the scores test and I was advised that omnibus test of
model coefficients is the likelihood ratio chi-square, or G, statistic, as
described in Hosmer and Lemeshow; and that the -2 Log likelihood in the
Model Summary table in SPSS is the D statistic described by Hosmer and
Lemeshow.

It has been suggested that running your analysis in NomReg should produce
the same results as Logreg as well as allowing you to test each of the
coefficients using the likelihood ratio test.



Bob