Date: Tue, 15 Apr 1997 16:57:21 GMT
Reply-To: Jodi Elliott <jelliott@RS1.TCS.TULANE.EDU>
Sender: "SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU>
From: Jodi Elliott <jelliott@RS1.TCS.TULANE.EDU>
Organization: Tulane University, New Orleans, LA
Subject: Logistic Regression
I have run a logistic regression analysis and need to report the results
in table format. Looking at examples in the published literature, I see
that several people have reported b (clear enough), s.e. (also clear),
chi-square, df, and something called "Pseudo R^2." Each predictor has a
b and s.e. reported, along with "Unique Pseudo R^2," then there is a
"Pseudo R^2" at the bottom of the table, after the chi-square.
My SPSS output provides -2 LL for both constant-only model and for full
model, a Goodness of Fit value, Model Chi-Square along with df and significance
level, Improvement of full model over constant-only model, B, S.E., Wald,
df, Sig (of Wald statistic, I believe), R, and Exp(B) which I have heard
from this group is an odds ratio.
Can someone tell me what, if anything, in my output corresponds to this
"Pseudo R^2" that people seem to be reporting out? Do I need to specify
something in my syntax that I'm currently missing? Thanks in advance for
any help you can offer.