home *** CD-ROM | disk | FTP | other *** search
- Comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
- Path: sparky!uunet!uvaarpa!darwin.sura.net!paladin.american.edu!auvm!CUMC.CORNELL.EDU!TOTHAV
- Posted-Date: Wed, 27 Jan 1993 19:17:23 -0500 (EST)
- Mime-Version: 1.0
- Content-Type: TEXT/PLAIN; charset=US-ASCII
- Message-ID: <Pine.3.05.9301271923.A704-a100000@cumc.cornell.edu>
- Newsgroups: bit.listserv.sas-l
- Date: Wed, 27 Jan 1993 19:17:23 -0500
- Reply-To: Arpad Vincent Toth <tothav@CUMC.CORNELL.EDU>
- Sender: "SAS(r) Discussion" <SAS-L@UGA.BITNET>
- From: Arpad Vincent Toth <tothav@CUMC.CORNELL.EDU>
- Subject: PROC CATMOD on PC 6.03
- Comments: To: sas-l%uga.BITNET@cornellc.cit.cornell.edu
- Lines: 24
-
- I'm new to the group and my first question is to anyone who has
- experience using this PROC to look at categorical data. My example is
- from Fienberg, ...Cross-Classified Cat. Data...Table 6.1
- My code:
- data seed; input type $ treatmnt $ outcome $ numcell @; cards;
- t1 h d 41
- t1 h za 59
- t1 l d 11 /*t1=long leaf/ t2=slash*/
- t1 l za 89 /*treatmnt h=high l=low*/
- t2 h d 12 /*outcome d=dead za=alive*/
- t2 h za 88
- t2 l d 5
- t2 l za 95;
- proc catmod data=seed;
- weight numcell;
- model type * treatmnt * outcome=_response_/nogls noparm noresponse;
- loglin type|treatmnt|outcome @2; run;
-
- SAS/STAT User's Guide (REL 6.03) gives examples but doesn't explain
- it very well. Any suggestions? I'm particularly interested in the
- log linear model, looking at different interactions. The chi square
- gives me a start but I want to look deeper into the data.
-
- Thanks in advance for any tips/suggestions.
-