Re the point raised by John Nelder via Dave Stephens: at a much more
elementary level, it seems that nowadays a lot of users out there are
relying on a well-known spreadsheet package, which is supplied as one
of a set of facilities together with w/p and other things, often
provided as part of the deal with a new PC. This software can
produce graphs and do a few calculations, but is really an
inadequate tool for any serious statistical analysis. Though I have
no proof of this, I'm sure that papers in which all the statistical
analysis has been performed using this package must have gone right
along the publication chain into journals, including ones that are
held in high regard. Shouldn't this be a matter of still greater
concern to us?
Robert Newcombe.
........................................................
Dr. Robert G. Newcombe
Department of Medical Computing and Statistics
University of Wales College of Medicine
Heath Park
Cardiff CF4 4XN
UK.
Email: [log in to unmask]
Telephone: 01222 742329 / 742311
Fax: 01222 743664
........................................................
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|