Re: Non-Linear Trends in contingency tables

Article View: bit.listserv.stat-l

Article #3627

Back to bit.listserv.stat-l Previous Next

#3627

From: ALLAN ROSEN (412) 624-0418
Date: Fri, 10 Jan 1992 01:45

22 lines
1931 bytes

Ping-Wu Li gave me a reference for my question:

  (briefly) For a RxC contingency table with ordinal classifications,
   is the difference between
      the chi-square test for general association and
      the chi-square test for correlation
   a chi-square test for non-linear departures?

The answer is Yes.  Ping-Wu Li refers to B.S Everitt's, 'The analysis of
contingency tables'.  Everitt (p51-56) gives an example with a 2xC contingency
table but notes the method is applicable to RxC tables, as well.

Ping-Wu Li notes that log-linear models with orthogonal contrasts could
also be used to test for non-linearity (CATMOD in SAS or LOGLINEAR in SPSS).
I was specifically interested in trying to use the General Mantel Score
statistics provided by FREQ in SAS.
--------------------------------------------------------------------------
Allan Rosen                              Internet: ALZADR@vms.cis.pitt.edu
Epidemiology Data Center                 Bitnet:   ALZADR@pittvax.bitnet
127 Parran Hall, 130 DeSoto Street       Fax:   (412) 624-3775
Pittsburgh, PA  15261                    Phone: (412) 624-0418
--------------------------------------------------------------------------

Technical Details

Message-ID: <01GF4J6TQU3OGH316O@vms.cis.pitt.edu>

Path: arc1-ca.pugleaf.net!sparky!uunet!europa.asd.contel.com!paladin.american.edu!auvm!PITTVMS.BITNET!ALZADR

Show Full Headers

Complete Article Headers

X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: f5efb,fe36153c6be2ccab
X-Google-Attributes: gidf5efb,public
X-Google-ArrivalTime: 1992-01-09 17:59:29 PST
Comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
Path: arc1-ca.pugleaf.net!sparky!uunet!europa.asd.contel.com!paladin.american.edu!auvm!PITTVMS.BITNET!ALZADR
X-Envelope-to: STAT-L@MCGILL1.BITNET
X-VMS-To: IN%"STAT-L@MCGILL1.BITNET"
X-VMS-Cc: ALZADR
Message-ID: <01GF4J6TQU3OGH316O@vms.cis.pitt.edu>
Newsgroups: bit.listserv.stat-l
Date:         Thu, 9 Jan 1992 20:45:00 EST
Sender:       "STATISTICAL CONSULTING" <STAT-L@MCGILL1.BITNET>
From:         "ALLAN ROSEN (412) 624-0418" <ALZADR@PITTVMS.BITNET>
Subject:      Re: Non-Linear Trends in contingency tables