home *** CD-ROM | disk | FTP | other *** search
- Comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
- Path: sparky!uunet!paladin.american.edu!auvm!CEHIEC1.EM.CDC.GOV!MJK1
- Encoding: 29 TEXT
- X-Mailer: Microsoft Mail V3.0 (beta-2)
- Message-ID: <2AA1A0CE@router.em.cdc.gov>
- Newsgroups: bit.listserv.stat-l
- Date: Mon, 31 Aug 1992 08:27:00 EST
- Sender: "STATISTICAL CONSULTING" <STAT-L@MCGILL1.BITNET>
- From: mjk1@CEHIEC1.EM.CDC.GOV
- Subject: Info on package called SUDAN?
- X-To: STAT-L@VM1.MCGILL.CA
- Lines: 27
-
- The package you are referring to is called SUDAAN (SUrvey DAta ANalysis). I
- believe you can get a copy of this by contacting Dr. Babu Shah, RTI, PO Box
- 12194, 3040 Cornwallis Rd., RTP, NC 27709 (address looks funny but this is
- all I can seem to locate in my notes) .
-
- I have used this package to run logistic regression models that account for a
- complex survey design. In addition to weighting each respondent in my sample
- for his or her probability of selection and other factors associated with the
- complex design of the survey, I did include a post-stratification component.
- However, had it not been necessary to account for the complex survey, I would
- have been able to perform my analyses in SAS and still use differing weights
- for each sample member.
-
- I have not looked into the use of this package for survival analysis.
- However, since SAS runs logistic regression with different weights applied,
- is it possible that it will do the same ina survuval analysis? The reason I
- ask is that I have found the documentation for SUDAAN to be confusing at
- times. In addition, I was under the impression that it's use was only
- necessary when there was a need to account for a complex survey design. In
- my situation, it was the estimated standard errors that differed (SAS vs
- SUDAAN) - my point estimates were identical.
-
- Marcie-jo Kresnow CDC (404)488-4656
-
-
- However, I will add that Dr. Shah has been available and very
- responsive, by phone, to my questions.
-