--- Jim Tantillo <[log in to unmask]> wrote:
> Now, not knowing the specifics of the animal testing procedures (and
> eagerly looking forward to your summary of it) for phthalate toxicity, I
> would still hazard the guess that the various tests were a lot like
> other
> such tests: extreme dosages of phthalate (10x or 100x or even 1000x
> more
> than what humans might encounter) were forced upon laboratory animals,
> with
> the results you report (testicular atrophy, etc.). I have no doubt this
> is
> the case. But then we still have to ask, what is the significance of
> those
> tests for inspiring the fear and concern over human children ingesting
> phthalates?
Here we have a crude version of Bayesian inference. Jim has probably read
about animal testing for other compounds and noted a similarity in several
cases. Using this information he has made the inference that the animal
testing for phthalates is similar and is waiting for John Foster to
provide him with new information. Once he has this new information Jim
will likely revise his initial position. Of course this is all ad hoc,
and an actual Bayesian approach would be much more rigorous, but you can
see how the mechanism works.
Another nice thing about the Bayesian approach is the prior selection.
Many see this as a drawback, but I think it is a strong point. The prior
is a probability distribution over the parameter of interest. Now, if we
want to be cautious we can select a prior that reflects this caution.
Also, if we are uncertain we can select a prior that has a large variance.
Then as new data comes available the prior is revised utilizing Bayes
Theorem and the new distribution, called a posterior, will change to
reflect this new information. For example the variance will usually
decrease indicating that as we gain new information we become more
certain.
In a sense, we can capture the much of the policy effects of the PP by
selecting a suitable prior and then letting new data inform us if we were
overly cautious or not.
> Don't get me wrong . . . maybe you know something that I don't, that
> phthalates in baby teethers pose a real risk--I'd like to see the
> evidence.
Actually, I don't even think the precautionary principle needs evidence,
just merely the possibility of harm. Further, there is no way in which to
get further evidence as whatever would generate the evidence would be
prohibited.
Steve
__________________________________________________
Do You Yahoo!?
Send instant messages & get email alerts with Yahoo! Messenger.
http://im.yahoo.com/
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|