Date:         Tue, 23 Jan 2007 20:39:07 +0000
Reply-To:     toby dunn <tobydunn@HOTMAIL.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         toby dunn <tobydunn@HOTMAIL.COM>
Subject:      Re: Creating a new group id by combining two other group id's via
              an array?
Comments: To: joeparts@GMAIL.COM
In-Reply-To:  <1169584256.393938.324000@m58g2000cwm.googlegroups.com>
Content-Type: text/plain; format=flowed

Why break them into arbitrary groups, seems to me like this would make for a
stats nightmare.   build, training and validation data set there are better
ways to do this.



Toby Dunn

To sensible men, every day is a day of reckoning. ~John W. Gardner

The important thing is this: To be able at any moment to sacrifice that
which we are for what we could become. ~Charles DuBois

Don't get your knickers in a knot. Nothing is solved and it just makes you
walk funny. ~Kathryn Carpenter






From: Joep <joeparts@GMAIL.COM>
Reply-To: Joep <joeparts@GMAIL.COM>
To: SAS-L@LISTSERV.UGA.EDU
Subject: Re: Creating a new group id by combining two other group id's via
            an array?
Date: Tue, 23 Jan 2007 12:30:56 -0800

I have received a request to further specify my question:

Well, the basic idea is the create separate datasets that I will
analyze in more detail. After I have selected about 20 important
datasets, I will use them to:
- calculate mean scores (correlations, sample sizes etc) for each
dataset
- estimate a multilevel model to predict certain values (proc mixed)
for each dataset.

So I will use these groups quite a lot. (Another solution would thus be
to think of simple way to create multiple datasets)

Hope this helps.
Joep

_________________________________________________________________
Laugh, share and connect with Windows Live Messenger
http://clk.atdmt.com/MSN/go/msnnkwme0020000001msn/direct/01/?href=http://imagine-msn.com/messenger/launch80/default.aspx?locale=en-us&source=hmtagline