Date: Fri, 9 Jun 2006 18:35:56 -0400
Reply-To: Kevin Roland Viel <kviel@EMORY.EDU>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Kevin Roland Viel <kviel@EMORY.EDU>
Subject: Re: additive effects: contrast?
Content-Type: TEXT/PLAIN; charset=US-ASCII
On Fri, 9 Jun 2006, Swank, Paul R wrote:
> have you tested for deviations from linearity? If so and there are none,
> then you can say the indication is that the relation is linear. Then the
> difference is b between each level and 2b between -1 and 1.
By "deviations from linearity" do you mean an alternative to the -1, 0, 1
coding? If so, I have not. It is the response we expect theoretically
and the coding I wish to test.
> Paul R. Swank, Ph.D.
> Professor, Developmental Pediatrics
> Director of Research, Center for Improving the Readiness of Children for
> Learning and Education (C.I.R.C.L.E.)
> Medical School
> UT Health Science Center at Houston
> -----Original Message-----
> From: SAS(r) Discussion [mailto:SAS-L@LISTSERV.UGA.EDU] On Behalf Of
> Kevin Roland Viel
> Sent: Friday, June 09, 2006 4:24 PM
> To: SAS-L@LISTSERV.UGA.EDU
> Subject: additive effects: contrast?
> In a multiple linear regression analysis, I have coded a covariate -1,
> 0, and 1. Supposed the estimate is a (p=b).
> The interpretation is that that the mean of my outcome in the group
> with 0 is a units higher than the mean in the group with -1. The same
> is true for the group with 1 compared to the group with 0. For either,
> I can say that p=b.
> The mean in the group with 1 is 2a higher than the mean in the group
> with -1. It seems to me that I cannot say that p=b, in this case.
> Would it be best to use a CONTRAST statement in this situation.
> Kevin Viel
> Department of Epidemiology
> Rollins School of Public Health
> Emory University
> Atlanta, GA 30322
Department of Epidemiology
Rollins School of Public Health
Atlanta, GA 30322