Neil Baylis wrote:

>>> If 220 is somehow more dangerous than 110, then there will be
>>> statistics (i.e., excess death rate from electrocution) to show that.

>> Statistics is not that easy. The death rate would be influenced by many
>> other factors, some of which might be more important than the voltage.

> Well then,  what fact about the world would constitute evidence that high
> voltage is better (or worse) than low voltage? Where would you look, if
> you don't look at death rate from electrocution?

I think death rate from electrocution would be a suitable measure. But you
can't just take different rates associated with different voltages and
compare them and then conclude what's safer.

The problem with applying science (Popper style and similar) to humans is
that we are somewhat reluctant to /really/ do it... run experiments, reduce
variables, all that kind of thing we do to dead matter (and some degree to
animals). 

So the proper (Popper) way would be to take a large enough sample of the
human species and put it into a controlled environment. Create that
environment so that the most common variables besides the voltage that
influence the accidental electrocution rate are the same for all HUTs. Then
leave the sample long enough in that environment to get statistically
relevant data. 

Just kidding... halfway, at least. Anyway, you'd have to analyze (analyse?)
the known statistics and try to create an image of the rates and how they
relate to the different factors that might influence them and their
statistics. Maybe you can get a trend out of it, but you surely need a
sample size of measurements (that is, independent statistics) that is
bigger than just two. This is how generally the sciences (or "sciences")
related to humans work, for lack of proper experimenting.

You also could go by personal preference. Of course this means that you'd
have to create a large enough body of experience with either, in different
situations. If we really get reborn, and if you can keep this up for a few
life cycles, you actually might get somewhere with this.

Gerhard

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist