Athel Cornish-Bowden wrote:
> >> I should be interested to know what readers of this news group think about
> >> entropy-enthalpy compensation,
> >
> >A name that springs to mind is that of Otto Exner, who has diligently
> >debunked several myths of this sort as statistical artifacts.
>> You are right. His contribution was one of the most important. I have read
> his papers on this (some of them, anyway) and have been very surprised that
> the subject didn't die a natural death after he pointed out the problems.
Well, we are a stubborn bunch. Sometimes we just don't want to know...
Why complicate a splendidly funded research project with annoying
statistical facts? I must say, the "ancients" occasionaly surprise me
with their level of rigor; why, they even knew calculus! Has anyone
looked at the awesome papers by Guldberg and Waage from the 1870's?
Impressive reading, I must say.
Incidentally, speaking of statistical artifacts, my particular 'beef' is
how we report results of nonlinear regression analyses in biochemical
literature. M. L. Johnson wrote several articles on the use of
least-squares regression,
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
@article { John9215,
author = "M. L. Johnson",
title = "Why, when, and how biochemists should use least squares",
journal = "Anal. Biochem.",
volume = "206", pages = "215-225",
year = "1992" }
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
@article { John9201,
author = "M. L. Johnson and L. M. Faunt",
title = "Parameter estimation by least-squares methods",
journal = "Meth. Enzymol.",
volume = "210", pages = "1-37",
year = "1992" }
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
@article { John8501,
author = "M. L. Johnson and S. G. Frasier",
title = "Nonlinear least-squares analysis",
journal = "Meth. Enzymol.",
volume = "117", pages = "301-342",
year = "1985" }
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
@article { John9401,
author = "M. L. Johnson",
title = "Use of least-squares techniques in biochemistry",
journal = "Meth. Enzymol.",
volume = "240", pages = "1-23",
year = "1994" }
and most people -still- use 'formal standard errors' of fitting
parameters, instead of full confidence intervals! This is actually
quite absurd, but I'd venture to say it'll go on forever.
The Exner references are not in my BibTeX database, but I -will- look
them up in my paper files...
-- Petr
_____________________________________________________________
Petr Kuzmic Ph.D. * BioKin Ltd. * Madison, WI 53708-8336, USA
pkuzmic at biokin.com * http://www.biokin.com * 608.256.1269 fax
