LISTSERV at the University of Georgia
Menubar Imagemap
Home Browse Manage Request Manuals Register
Previous messageNext messagePrevious in topicNext in topicPrevious by same authorNext by same authorPrevious page (January 2007, week 3)Back to main SAS-L pageJoin or leave SAS-L (or change settings)ReplyPost a new messageSearchProportional fontNon-proportional font
Date:         Thu, 18 Jan 2007 02:09:08 -0500
Reply-To:     Ankur Arora <ankur.arora@NAGARRO.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Ankur Arora <ankur.arora@NAGARRO.COM>
Subject:      Re: problem of parametric trend test
Comments: To: "David L. Cassell" <davidlcassell@MSN.COM>
Content-Type: text/plain; charset=ISO-8859-1

Hi David,

I have a dataset like this

group Independent variable

1 123 1 124 1 122 2 123 2 111 2 123 3 123 3 123 3 344 4 123 4 234

so my class the variable is group.

1) I had one option with proc logistics

proc logistic; model GROUP(order=internal ref="1")=var to be analysed / link=glogit risklimits; run;

1 :is the control group so basically finding trend by comparing with the control group.

2) option is to use proc reg with contrast statement but here we are not specifying any control group

how do i generate the contrast coeficient acording to no of groups because i need to program a macro where no of groups can change so i need to generate contrast statement at run time.

I am confused how to calculate the trend. could you throw light on how to do parametric trend statistically.

so how do I set the parametric trend.

Thanks Ankur

Original Message: ----------------- From: David L Cassell davidlcassell@MSN.COM Date: Wed, 17 Jan 2007 22:27:36 -0800 To: SAS-L@LISTSERV.UGA.EDU Subject: Re: problem of parametric trend test

ankur.arora@NAGARRO.COM wrote: > >Hi All, > > >I want to perform parametric trend test in sas. > >Can anybody advise how to perform the parametric trend test in sas. > >Thanks in advance. > >Thanks >Ankur

There is no single "the parametric trend test", unfortunately. And if there were only one, it probably wouldn't work for most data sets. :-(

So you need to describe your data and meta-data in a lot more detail, with some discussion on the overall goal here, so we have some idea of what kind of test is likely to be reasonable.

Then you would end up having to check the assumptions of whatever test gets recommended, and decide if the test can even be applied to your data.

HTH, David -- David L. Cassell mathematical statistician Design Pathways 3115 NW Norwood Pl. Corvallis OR 97330

_________________________________________________________________ FREE online classifieds from Windows Live Expo – buy and sell with people you know http://clk.atdmt.com/MSN/go/msnnkwex0010000001msn/direct/01/? href=http://expo.live.com?s_cid=Hotmail_tagline_12/06


Back to: Top of message | Previous page | Main SAS-L page