My paper on correlations

[From Richard Kennaway (970808.1924 BST)]

A few months ago I posted here about a paper on the uselessness of
correlational data for making individual predictions. I'd like to get it
published somewhere prominent, and I've rewritten it to fit into the form
expected of a paper to "Science" or "Nature". The new version is available
on the web as a downloadable PostScript file. The URL is
http://www.sys.uea.ac.uk/~jrk/distribution/correlationinfo.ps
(uncompressed, 507k) or
http://www.sys.uea.ac.uk/~jrk/distribution/correlationinfo.ps.gz
(compressed with gzip, 109k). For those with ftp but not web access, I've
also put those files in the directory
ftp://ftp.sys.uea.ac.uk/pub/kennaway/drafts, but there may be problems with
our ftp site.

I've moved all the mathematical proofs into an appendix, which would not be
intended to be part of the published version. I've made various other
minor changes.

I'd appreciate any comments, especially from people with more experience of
those journals and more familiar with what is most likely to succeed (in
terms of style -- can't do anything about the message :-)). It could
probably do with more references, and more recent ones, to examples of low
correlations in the literature. Opening with a reference to McPhail 1971
hardly gives an impression of novelty.

-- Richard Kennaway, jrk@sys.uea.ac.uk, http://www.sys.uea.ac.uk/~jrk/
   School of Information Systems, Univ. of East Anglia, Norwich, U.K.