R help - Dec 2005 - Problem with understanding output of Cox model

Hi all,
 
I am using a 'tricked' Cox Hazard regression model for discrete choice
habitat modelling.
However, I'm having a hard time understanding the meaning of the first
line the following part of the summary() output:
 
Rsquare= 0.307 (max possible= 0.475 )
Likelihood ratio test= 91.8 on 12 df, p=2.23e-14
Wald test = 26.3 on 12 df, p=0.00977
Score (logrank) test = 58.6 on 12 df, p=4.03e-08
 
Does anyone know how I can read the 'max possible' R-square? What does
it signify?
Is it some kind of quasi-correlation; in which case I read it that I
have explained 65% (i.e., 0.307/0.475) of the variation?
 
I hope someone can help on this.
Thanks in advance,
 
Cheers Roel
 
 
Roel May
Norwegian Wolverine Project
Norwegian Institute for Nature Research (NINA)
Tungasletta 2, N-7485 Trondheim, Norway
Tlf. +47 73 80 14 65, Mob. +47 95 78 59 95
Email roel.may@nina.no <mailto:roel.may@nina.no> 
Internett www.nina.no <http://www.nina.no/> , www.jerv.info
<http://www.jerv.info/>  
 

	[[alternative HTML version deleted]]

On Tue, 13 Dec 2005, May, Roel wrote:
> Hi all,
>
> I am using a 'tricked' Cox Hazard regression model for discrete
choice
> habitat modelling.
> However, I'm having a hard time understanding the meaning of the first
> line the following part of the summary() output:
>
> Rsquare= 0.307 (max possible= 0.475 )
> Likelihood ratio test= 91.8 on 12 df, p=2.23e-14
> Wald test = 26.3 on 12 df, p=0.00977
> Score (logrank) test = 58.6 on 12 df, p=4.03e-08
>
> Does anyone know how I can read the 'max possible' R-square? What
does
> it signify?
It doesn't signify anything very useful.  It is the proportional reduction 
in deviance.

 	-thomas