home *** CD-ROM | disk | FTP | other *** search
/ NetNews Usenet Archive 1992 #19 / NN_1992_19.iso / spool / bit / listserv / statl / 1413 < prev    next >
Encoding:
Text File  |  1992-08-27  |  1.1 KB  |  33 lines
  1. Comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
  2. Path: sparky!uunet!gatech!paladin.american.edu!auvm!UNCA.BITNET!WEST
  3. X-Organization: University of North Carolina at Asheville
  4. X-Envelope-to: stat-l@mcgill1.bitnet
  5. X-VMS-To: IN%"stat-l@mcgill1.bitnet"
  6. Message-ID: <01GO2QVCAJDS0009Y2@UNCA.BITNET>
  7. Newsgroups: bit.listserv.stat-l
  8. Date:         Thu, 27 Aug 1992 12:20:00 EST
  9. Sender:       "STATISTICAL CONSULTING" <STAT-L@MCGILL1.BITNET>
  10. From:         WEST@UNCA.BITNET
  11. Subject:      Post-stratification of surveys question
  12. Lines: 19
  13.  
  14. A Survey Question for Stat-L:
  15.  
  16.         I am doing some work that involves the post-stratification
  17. of a statewide survey.  Essentially, we tend to undersample certain
  18. ethnic, age, and educational groups, and so we post-stratify to
  19. correct for this design effect.
  20.  
  21.         Cochran (1975) gives a procedure for determining the relative
  22. efficiency of post-stratification when compared to the non-stratified
  23. survey results; does anyone know of a comparable method for comparing
  24. two different post-stratification schemes?
  25.  
  26.         Thanks in advance..
  27.  
  28. Mark West
  29. WEST@UNCA.EDU / WEST@UNCA.BITNET
  30. Dept. Mass Comm.
  31. UNC-Asheville
  32. Asheville, NC  28804
  33.