home *** CD-ROM | disk | FTP | other *** search
/ NetNews Usenet Archive 1992 #18 / NN_1992_18.iso / spool / bit / listserv / statl / 1340 < prev    next >
Encoding:
Text File  |  1992-08-20  |  1.5 KB  |  39 lines

  1. Comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
  2. Path: sparky!uunet!uvaarpa!darwin.sura.net!paladin.american.edu!auvm!NIOSHE2.EM.CDC.GOV!DKW4
  3. Return-Receipt-To: <dkw4@NIOSHE2.EM.CDC.GOV>
  4. Encoding: 27 TEXT
  5. X-Mailer: Microsoft Mail V3.0 (beta-2)
  6. Message-ID: <2A939339@router.em.cdc.gov>
  7. Newsgroups: bit.listserv.stat-l
  8. Date:         Thu, 20 Aug 1992 16:40:00 EST
  9. Sender:       "STATISTICAL CONSULTING" <STAT-L@MCGILL1.BITNET>
  10. From:         dkw4@NIOSHE2.EM.CDC.GOV
  11. Subject:      Correction for Shapiro-Wilk test
  12. X-To:         STAT-L%MCGILL1.BITNET@CUNYVM.CUNY.EDU
  13. Lines: 24
  14.  
  15. Question?  Has anyone heard of a correction factor for the Shapiro-Wilk test
  16. (for normality) produced by SAS's PROC UNIVARIATE?   If you have heard of
  17. such a correction factor, do you think it really helps?
  18.  
  19. That is, I seem to find the S-W test to lack power for small sample sizes and
  20. to almost always reject the null hypothesis (of normality) for large sample
  21. sizes.  Translation:  I find it to be a worthless test most of the time.
  22.  
  23. However, a colleague (who was quoting me....gulp) was challenged by another
  24. person who said Shapiro-Wilk worked just fine with the correction factor.
  25.  
  26. So, does anyone know anything about this?  Thanks much for any help.
  27.  
  28. Deanna
  29.  
  30. *-----------------------------------------------------------------*
  31. Deanna Wild
  32. Statistician
  33. National Institute for Occupational Safety and Health
  34. 4676 Columbia Parkway  MS: R-4
  35. Cincinnati, Ohio  45226
  36.  
  37. INTERNET: DKW4@NIOSHE2.EM.CDC.GOV
  38. *-----------------------------------------------------------------*
  39.