home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.math.stat
- Path: sparky!uunet!paladin.american.edu!howland.reston.ans.net!spool.mu.edu!agate!dog.ee.lbl.gov!news!manta!herman
- From: herman@nosc.mil (John W. Herman)
- Subject: Re: Robust correlation
- Message-ID: <herman.727633991@phage>
- Keywords: robust, correlation
- Sender: usenet@nosc.mil (Network News)
- Organization: Naval Ocean Systems Center, San Diego, CA
- References: <1993Jan21.012621.21074@jyu.fi>
- Date: Thu, 21 Jan 1993 16:33:11 GMT
- Lines: 17
-
- vms@jyu.fi (Veli Matti Suppola) writes:
-
- >A few days ago I posted a question to this newsgroup. The question
- >was about robust correlation coefficients.
- >I got a few answers but I do know, that many of You know something
- >about this subject.
- >Thanks to all of You, who answered my question.
- >I am very grateful.
- >So, if You know something about robust correlation coefficients,
- >please let me know. I mean books, articles and so on.
-
- I'm not sure what you're trying to find. The Rank Correlation
- Coeffecient is a robust approximation to the true correlation. If you
- have symmetrical distributions, you can also look at sign matching. I'm
- sure that both of these have been mentioned to you by other people.
- Maybe a description of your problem would help elicit a better
- response.
-