SPSS and Survival Analysis :(

Thread View: bit.listserv.spssx-l

3 messages

[T] Back to Threads [V] Tree View [G] Group Articles

3 total messages Started by Phillip Phan Fri, 16 Jul 1993 02:33

#287

Author: Phillip Phan
Date: Fri, 16 Jul 1993 02:33

31 lines
2311 bytes

Hi all,
	Does anyone out there know a reasonable workaround to using
	SPSS Survival with continous covariates?  SPSS forces you to
	convert ratio measures to interval ones and there are major
	problems with that approach theoretically (loss of info, 
	defensibility of the break points chosen, etc..)  I know that
	BMDP implements the Cox Proportional Model, which is the
	recommended approach for ratio covariates, beautifully; but
	I'm fully invested in SPSS and don't wish to buy and learn a
	new environment just for this project.  Is there a solution
	or is this one of those famous walls that you simply have to
	live with as a tradeoff for user friendliness?  I thought of
	computing the harzard rate and then exporting it into loglinear
	as a dependent.  Any problems with this approach?  Please help.
	Thanks.
Phil Phan
-- 
================================================================================
	     All opinions are mine and are not necessarily the views
		     of York University or it's affiliates
 *********************************** 
|         Phillip H. Phan	    |    __________________________________
|         York University	    |   | 				   |
|	   FAS - POLICY 	    |   | "Let's kick some elite ass..."   |
|      North York, ON M3J 1P3	    |   |				   |
|	       CANADA		    |   |  				   |8888
|           416-736-5087	    | 	|	            -- Al Bundy	   |8888
|      as000047@orion.yorku.ca	    |   |__________________________________|8888
 ***********************************         88888888888888888888888888888888888
					     88888888888888888888888888888888888
					     88888888888888888888888888888888888

Re: SPSS and Survival Analysis :(

#288

Author: John Douglas
Date: Fri, 16 Jul 1993 12:36

25 lines
2096 bytes

>Hi all,
>	Does anyone out there know a reasonable workaround to using
>	SPSS Survival with continous covariates?  SPSS forces you to
>	convert ratio measures to interval ones and there are major
>	problems with that approach theoretically (loss of info, 
>	defensibility of the break points chosen, etc..)  I know that
>	BMDP implements the Cox Proportional Model, which is the
>	recommended approach for ratio covariates, beautifully; but
>	I'm fully invested in SPSS and don't wish to buy and learn a
>	new environment just for this project.  Is there a solution
>	or is this one of those famous walls that you simply have to
>	live with as a tradeoff for user friendliness?  I thought of
>	computing the harzard rate and then exporting it into loglinear
>	as a dependent.  Any problems with this approach?  Please help.

I had similar problems, now I use SAS and rarely BMDP.

John
                                GopherStuffer
+-------------------------------------+-------------------------------+
| John Douglas, Information Manager   | Internet jd@ssru.city.ac.uk   |
| Social Statistics Research Unit     | JANET    jd@uk.ac.city.ssru   |
| City University, Northampton Square | Voice    +44-71-477-8490/8481 |
| London EC1V 0HB, UK                 | Fax      +44-71-477-8583      |
+ ------------------------------------+-------------------------------+

Re: SPSS and Survival Analysis :(

#289

Author: David Nichols
Date: Fri, 16 Jul 1993 18:44

27 lines
2267 bytes

>>	Does anyone out there know a reasonable workaround to using
>>	SPSS Survival with continous covariates?  SPSS forces you to
>>	convert ratio measures to interval ones and there are major
>>	problems with that approach theoretically (loss of info, 
>>	defensibility of the break points chosen, etc..)  I know that
>>	BMDP implements the Cox Proportional Model, which is the
>>	recommended approach for ratio covariates, beautifully; but
>>	I'm fully invested in SPSS and don't wish to buy and learn a
>>	new environment just for this project.  Is there a solution
>>	or is this one of those famous walls that you simply have to
>>	live with as a tradeoff for user friendliness?  I thought of
>>	computing the harzard rate and then exporting it into loglinear
>>	as a dependent.  Any problems with this approach?  Please help.

SPSS implemented Cox's proportional hazards regression model, including
both fixed and time dependent covariates, in version 5. It is available
on releases 5 and 6 (now available) for Windows. It is not in PC+. It
will be on mainframe and Unix releases numbered 5 and above when they
come out (the only one currently out to the best of my knowledge is
Motif for the RS/6000). I think you'll find our implementation of COXREG
stands up with that of either SAS or BMDP.

--
 David Nichols        Senior Statistical Support Specialist         SPSS, Inc.
 Phone: (312) 329-3684     Internet:  nichols@spss.com     Fax: (312) 329-3657
*******************************************************************************
 Any correlation between my views and those of SPSS is strictly due to chance.

Thread Navigation

This is a paginated view of messages in the thread with full content displayed inline.

Messages are displayed in chronological order, with the original post highlighted in green.

Use pagination controls to navigate through all messages in large threads.

Back to All Threads