[ QUOTE ]
ANY standard is better than no standard at all.
[/ QUOTE ]
The counter argument to this is that a certification implies some sort of
competence (usually), and having one that in fact doesn't really depend on
this would be worse than no certification at all, at least in that w/o any sort
of certification one would not be misled. I.e., if certification makes no
difference, if it only actually proves that someone jumped throught the
hoops of some process (but not that they have the skills implied), then
you're better off w/o it, leaving assessment of abilities to some other
method.
I surmise that in arriving at some level of testing to give certification
there was a line-drawing issue of trying to please eveybody, not wanting
to require college study, etc.. One way a certification scheme might go
to meet some differing needs is to complement some minimal demands
with additional components or grading. --levels of certification. E.g.,
the Ada programming language was born with requirements that
any certified language processor must support the entire language,
not merely a subset. Upon its revision for the 1995 standard, there was
demand to make the language more capable in a variety of areas (for
information systems, real-time, safety, numerics); but with concern to
not overburden every processor with all of these specialized capabilities,
they were included as Specialized Needs Annexes, which processors
could implement none, some, or part of. And whereas the designers
wanted to have certification for these annexes like that for the core
language (all or none), it was decided rather to allow partial support
to be tested & reported with the degree of support indicated in the
test report & certificate.
One might imagine that with such packaging & grading, a certification
for arborists might be able to be fuller & demanding, but still be able
to be passed with an acceptable minimum capability (and I note that
it is argued above that the current level is too low).
*knudeNoggin*