home *** CD-ROM | disk | FTP | other *** search
- Comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
- Path: sparky!uunet!uvaarpa!darwin.sura.net!paladin.american.edu!auvm!IRLEARN.UCD.IE!RCONROY
- Message-ID: <STAT-L%93012808242166@VM1.MCGILL.CA>
- Newsgroups: bit.listserv.stat-l
- Date: Thu, 28 Jan 1993 13:19:48 GMT
- Sender: STATISTICAL CONSULTING <STAT-L@MCGILL1.BITNET>
- From: Ronan M Conroy <RCONROY@IRLEARN.UCD.IE>
- Subject: Re: correlation coefficients
- In-Reply-To: Message of Wed, 27 Jan 1993 16:34:00 EST from <KIRISCI@PITTVMS>
- Lines: 17
-
- On Wed, 27 Jan 1993 16:34:00 EST <KIRISCI@PITTVMS> said:
- >
- >I have two questionnaries:one has items with 7-point scale and the other one
- >has
- > items with 3-point scale. I wonder if anyone suggest me how to calculate
- >correlation coefficients between individual items whose have different scales.
- >What can be the best way? or do I need to make some transformations?
- >I am looking forward to reading your comments.
- >thank you.
-
- Seems to me that this is a case where Kendall's tau should work better
- than Spearman's rho, since one of the scales has few values.
- But, apart from the old SPSS manual (V5?) has anyone a reference
- on when tau works better than rho?
-
-
- ronan
-