|Date: ||Thu, 12 Jul 2007 10:34:33 -0400|
|Reply-To: ||"V. Bourcier" <veronique.bourcier@LAPOSTE.NET>|
|Sender: ||"SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>|
|From: ||"V. Bourcier" <veronique.bourcier@LAPOSTE.NET>|
|Subject: ||Re: 'class' statement versus creating dummy variables.|
I'm taking the opportunity of this class vs. dummy variable discussing to
extend it to the lifetest procedure.
I've been said on a former study to get treatment values which follows one
another i.e. 0,1 would do, but also 5,6, while 2,5 would affect the hazard
ratio. This would be due to the absence of class variable.
1. Does this sound all correct to you? Based on so far replied on the
topic, a better understanding of the dummy concept would help me too. Am I
2. Would you say any further option need to be set?
3. Would you advice any further documentation on the topic?
Any reply should be helpful for any further work I may have to do.