home *** CD-ROM | disk | FTP | other *** search
/ NetNews Usenet Archive 1992 #19 / NN_1992_19.iso / spool / bit / listserv / statl / 1451 < prev    next >
Encoding:
Text File  |  1992-09-02  |  818 b   |  18 lines
  1. Comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
  2. Path: sparky!uunet!paladin.american.edu!auvm!UCLAMVS.BITNET!IARSPAL
  3. Message-ID: <STAT-L%92090215521725@VM1.MCGILL.CA>
  4. Newsgroups: bit.listserv.stat-l
  5. Date:         Wed, 2 Sep 1992 12:48:00 PDT
  6. Sender:       "STATISTICAL CONSULTING" <STAT-L@MCGILL1.BITNET>
  7. From:         "Peter A. Lachenbruch" <IARSPAL@UCLAMVS.BITNET>
  8. Subject:      unequal variances
  9. Lines: 7
  10.  
  11. There is another aspect to the unequal variance problem - that of power under
  12. alternatives. In a recent study, (Comm. Stat., 1991) I compared the size and
  13. power of the F, Kruskal-Wallis, and Normal Scores tests when variances were un-
  14. equal. The F test was badly hurt, but the KW and NS tests were relatively un-
  15. affected. I suspect that the rank transformation will work pretty well in most
  16. anova problems.
  17. Tony
  18.