Date: Tue, 3 Sep 1996 16:45:41 +0000
Reply-To: Dmitry Mesyanzhinov <dmes@WWW.LEERIC.LSU.EDU>
Sender: "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From: Dmitry Mesyanzhinov <dmes@WWW.LEERIC.LSU.EDU>
Subject: Measures of fit for LIFEREG
I would like to use PROC LIFEREG to
run Tobit model (left-censored data).
According to the SAS/SAT User's Guide,
the only available statistic for the measure
of fit is log-likelihood ratio (am I missing something?).
What is the proper way to access the goodness
of fit for LIFEREG models? By comparing to OLS estimates?
Thanks in advance.