Date:         Tue, 16 Mar 1999 13:48:15 -0500
Reply-To:     "Scott Compton, Ph.D." <scompton@ACPUB.DUKE.EDU>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU>
From:         "Scott Compton, Ph.D." <scompton@ACPUB.DUKE.EDU>
Subject:      loglinear parameters
Content-Type: multipart/alternative;

Hi Folks,
 
Can somebody explain to me why the different loglinear modules in SPSS create different parameter estimates. For instance, the parameter estimates in the output from GENLOG are different from those in LOGLINEAR and HILOGLINEAR. Is it simply due to the different algorithms used? Which ones should be used when reporting model parameter estimates or to calculate odds ratios.
 
Thanks in advance,
 
Scott Compton
Duke University
 
 
 
 
[text/html]