SAS GLM repeated-measures contrasts

Article View: bit.listserv.stat-l

Article #1012

Back to bit.listserv.stat-l Previous Next

#1012

From: mdspre
Date: Tue, 25 Jan 1994 19:32

19 lines
1551 bytes

Hi,
As usual, I'm having trouble with contrasts with repeated measures
designs in GLM in SAS (VMS version 6.02).  The built-in contrast
mechanisms, such as POLYNOMIAL, are OK for tests within one
repeated factor, but what do you do when you have a fully repeated
design with 3 factors, and you need to deal with complex contrasts (say
between levels of one factor, measured at a single level of another
factor)?  It must be possible to customize an appropriate M-matrix, but
this seems a lot of work to do by hand.  Is there some easy method to do
this?  Has anyone developed software to generate this sort of contrast --
something that lets you define your design in words, and allows you to
specify level comparisons?
Thanks in advance,
Elliot Sprecher, Ph.D.
Experimental Psychologist, Research Coordinator
Institute of Clinical Neurophysiology
Rambam Medical Center
Haifa, Israel
MDSPRE@VMSA.TECHNION.AC.IL

Technical Details

Message-ID: <1994Jan25.213221.1@vmsa.technion.ac.il>

Path: arc1-ca.pugleaf.net!gmd.de!dearn!barilvm!technion!discus.technion.ac.il!vmsa.technion.ac.il!mdspre

Show Full Headers

Complete Article Headers

X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: f5efb,733fa794ccf798e8,start
X-Google-Attributes: gidf5efb,public
X-Google-ArrivalTime: 1994-01-25 23:55:06 PST
Newsgroups: bit.listserv.stat-l
Path: arc1-ca.pugleaf.net!gmd.de!dearn!barilvm!technion!discus.technion.ac.il!vmsa.technion.ac.il!mdspre
From: mdspre@vmsa.technion.ac.il
Subject: SAS GLM repeated-measures contrasts
Message-ID: <1994Jan25.213221.1@vmsa.technion.ac.il>
Lines: 20
Sender: news@discus.technion.ac.il (News system)
Nntp-Posting-Host: vmsa.technion.ac.il
Organization: Technion, Israel Institute of Technology
Date: Tue, 25 Jan 1994 19:32:21 GMT