home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.math.stat
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!malgudi.oar.net!news.uakron.edu!news.csuohio.edu!vmcms.csuohio.edu!R0264
- From: R0264@vmcms.csuohio.edu
- Subject: Re: Measures of Multi-Modality
- Message-ID: <16B5EC4CB.R0264@vmcms.csuohio.edu>
- Sender: news@news.csuohio.edu (USENET News System)
- Organization: CSU
- References: <1jkb22INN8ob@pith.uoregon.edu>
- Date: Fri, 22 Jan 1993 18:59:25 GMT
- Lines: 18
-
- In article <1jkb22INN8ob@pith.uoregon.edu>
- joe@decoy.uoregon.edu (Joe St Sauver) writes:
-
- >
- >I've got a researcher who is looking for a measure of
- >multi-modality (or "bi-polarity," if you will) such
- >that unimodal data would have a small score, and bimodal
- >or trimodal or n-modal data would have a large score.
- >
- >I'm at a loss for what to recommend -- any suggestions?
- >
- >Thanks,
- >
- >Joe St Sauver (joe@oregon.uoregon.edu)
- The sample measure of kurtosis would be one approach, but might not be
- sufficient. It is 4th central moment divided by square of 2nd, if I
- remember correctly, and has value 3 for normal, and smaller values for
- platykurtic distributions.
-