home *** CD-ROM | disk | FTP | other *** search
- Comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
- Path: sparky!uunet!paladin.american.edu!auvm!UNHH.UNH.EDU!W_URBAN
- X-Vmsmail-To: @STAT-L
- Message-ID: <920722083615.202116e5@UNHH.UNH.EDU>
- Newsgroups: bit.listserv.stat-l
- Date: Wed, 22 Jul 1992 08:36:15 -0400
- Sender: "STATISTICAL CONSULTING" <STAT-L@MCGILL1.BITNET>
- From: W_URBAN@UNHH.UNH.EDU
- Subject: Multiple comparisons w/ repeated measures or split plot
- structures
- X-To: STAT-L@VM1.MCGILL.CA
- Lines: 12
-
- Under the new PROC MIXED in SAS, there is a /CL switch (confidence
- limits) for the LSMEANS statement. It would be interesting for
- someone to compare that output w/ that from SPSS.
-
- n.b. With unequal numbers of subjects per treatment, no two
- methods/packages that I have used have given the same F ratios.
-
- Willard Urban
- Biometrics
- University of New Hampshire
-
- will@unh.edu
-