Bill (and others),
Interesting perspective. I am a cynic too, but I take a different
angle. In my graduate-level biostatistics class here in Brazil, the
students have to read and analyse papers that are using the particular
analysis that we are working on that week. I am amazed at the bad
statistics, as you are. Regressions that are really correlations,
ANOVAS that should have been regressions, incomprehensible multivariate
analyses, complete failure to attend to the assumptions, hypotheses that
will be rejected by definition and not by compliance with (or not)
theory. The students all feel like, since they are reading most papers
in English, that they are the ones who don't understand, thinking that
all these published papers in many important journals must have gotten
published because they were well-written. I have to teach them that,
surprising though it may be, these papers have flaws from small to large.
HOWEVER, when well-done, statistics clear the confusion. After all,
psychologists show that people see patterns where they do not exist. I
would say that the "obvious" patterns do not necessarily need
statistics, and the self fulfilled prophecies do not either. However,
those kinds of results are boring and obvious. They were probably
predictable on general principles and the laws of physics.
The not-so-obvious results are the ones that are interesting, and those
are also the ones that need statistics to make sure that we are not
seeing patterns that do not exist. Without and idea of sampling error
and variance, our intuition over whether a pattern exists is
error-prone. The only way to control that error is statistics and
well-defined studies.
So, I say, we need to force the scientific and ecological community to
learn how to use the tools of the trade. Nobody need to be immersed in
statistics to understand the rules. But, if a person has not taste or
patience for statistics, then I suggest that they find a good
collaborator who knows statistics well. After all, in most of my
helping students develop projects, because of my statistical
understanding, I save them all time and frustration. The sample size
required to show a pattern is much easier to calculate with a knowledge
of statistics, for example.
Cheers,
Jim
Bill Silvert wrote:
>I didn't expect much agreement with my posting, and I'll just comment on two
>points that Roper raises, interspersed with his posting below:
>
>
>
=============================================
Consulto ECONCIÊNCIA
Ecologia, Conservação, Ciência e Consciência!
Consulting, specializing in Conservation
Research Methods, Analysis and Translations.
http://www.montanhaviva.org/ecosci/
=============================================
James J. Roper, Ph.D.
Caixa Postal 19034
81531-990 Curitiba, Paraná, Brasil
---------------------------------------------
E-mail: [EMAIL PROTECTED]
Phone/Fone/Teléfono: 55 41 33857249
celular: 55 41 99870543
e-fax: 1-206-202-0173 (in the USA)
---------------------------------------------
Ecologia e Conservação na UFPR
http://www.bio.ufpr.br/ecologia/
--
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.1.362 / Virus Database: 267.12.8/161 - Release Date: 3/11/2005
