home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky sci.skeptic:19268 rec.org.mensa:8106
- Path: sparky!uunet!mcsun!uknet!mucs!nessie!db.mcc.ac.uk!zlsiida
- From: zlsiida@fs1.mcc.ac.uk (dave budd)
- Newsgroups: sci.skeptic,rec.org.mensa
- Subject: Re: Astrology: Scientific Research
- Message-ID: <zlsiida.379@fs1.mcc.ac.uk>
- Date: 9 Nov 92 16:08:06 GMT
- References: <1992Oct14.163738.2878@lmpsbbs.comm.mot.com> <wiegand.719244521@lido16> <1992Oct27.020725.6275@Cadence.COM> <zlsiida.353@fs1.mcc.ac.uk> <1992Nov6.143346.960@msuinfo.cl.msu.edu>
- Sender: news@nessie.mcc.ac.uk (Usenet News System)
- Organization: Manchester Computing Centre
- Lines: 25
-
- In article <1992Nov6.143346.960@msuinfo.cl.msu.edu> banghartr@clvax1.cl.msu.edu (Rick Banghart, IPTV) writes:
- >In article <zlsiida.353@fs1.mcc.ac.uk>, zlsiida@fs1.mcc.ac.uk (dave budd) writes...
- >>there is an incredibly strong negative correlation between the soccer
- >>matches I predict as score draws and the ones that actually do turn out to
- >>be. but there is no causal link.
-
- >Pardon me as a speak from the dangerous place of "a little knowlege."
- >When I spoke of "strong statistical correlation" I assumed that this
- >was a correlation greater than that produced by chance. My hunch
- >is that your negative correlation between predicted match outcomes
- >and actual outcomes would not exceed that produced by chance.
- >If the negative correlation does exceed the chance level, then I
- >would begin to look for a reason. Again, I am not attributing
- >_meaning_, I am just concerned that we pay attention to the
- >facts, no matter how they may contradict our previously held ideas.
-
- Ahem, I exagerated: to make the point that correlation <> cause.
- Parts of my method have consistently negative correlation, parts have
- consistently positive correlation, parts are just inconsistent. Statistical
- significance of thses correlations is dubious. I've been investigating
- these stats for 5 years now and I don't want any more advice.....
-
-
- You don't have to work to be mad here email: d.budd@mcc.ac.uk
- Dave Budd, MCC, Oxford Rd, Manchester, England (44|0)61-275-6033
-