modelling

Peter Rauch peterr at SOCRATES.BERKELEY.EDU
Tue Sep 3 12:38:10 CDT 2002


On Tue, 3 Sep 2002, Anita F. Cholewa wrote:
> ...  Does anyone remember the incident in Oregon wherein a
> land manager planted lynx hair in an area where there was no
> scientific data to support lynx presence?  Perhaps his model
> said lynx should be in that area and so ...

Are you referring to this set of incidents, which reports on
what turned out to be a single-blind test, and had no intention
apparently to attempt to imply that lynx were present in the
areas from which the samples were purported to have come ?

Linkname: Lynx Hair Snagging Investigation:
     URL: http://www.ecosystem.org/tccp/supermatch/lynxreport.pdf

If so, I don't think this is an example of why one should beware
of either the foibles of modelling nor of the disreputable use
or other misuse of models. The lynx hair


There are a number of other "official" reports to be found on
the web as well. One states, for example, "The Protocol did not
provide procedures to submit hair samples collected outside the
survey to test the accuracy of laboratory results." While the
extraordinary measures taken by some hair-sampling employees may
be views as "unauthorized" and/or less than adequately-designed
"tests" of the system, and some suggest even "unethical", we
must ask why there was no protocol established to reliably test
the fidelity and accuracy of the _entire_ process, not just that
of the laboratory results.

I'm glad that these "reckless" employess ventured forth, risking
their jobs and reputations, and demonstrated --no matter whether
ineptly or not-- that there's plenty of room for improvements in
our protocols, as there is in our modelling enterprises.

Peter




More information about the Taxacom mailing list