"~^Johnny^~" wrote in message
...
On Sun, 12 Sep 2004 21:08:41 -0700, Robert Morein
wrote:

In article ,
wrote:

Sir TURTLE wrote:

Here is one for you. 5%RH or there about and 90ºF = pretty cool.
What do you think...

I'd say hot. So would ASHRAE, (Y = 1.79, with 67% of people
dissatisfied,
vs Y = 0, with 6% dissatisfied), based on 21,000 people around the
world,
who'd prefer adding some moisture to reduce the air temp, which
might come
from a swimming pool :-)

Imagine yourself in a 90 F office, in a dry shirt...

Still you living out of a book and do not understand the part %RH
plays
in the comfort levels.

I understand thoroughly, my good man. In this case, I'm afraid you are
the one lacking average empirical understanding, and an understanding
of the way surveys work.

I would have 90ºF and a 10%RH in my home before I would have 60ºF
with a 100%RH.

OK. Everyone's tastes are different. There's nothing wrong with your
misunderstanding average human behavior. It's a matter of surveys and
preferences, vs absolute science. But you must admit facts. It's hard
to deny the average tastes of 21,000 people around the world :-)

Nick


Do those 21,000 people from around the world reflect your tastes?

Here is where statistics fail.

which 21,000 people?

Is the target random and unbiased?
It is not!

Only paramaters will do.
Statistics are, statistically, on a bias. g
Their results are slanted. They can err on many facets (sides, if you
will).


It's not that the statistics fail. It's that they often get misused. But
if a competent engineer or scientist can figure out that failure by looking
at the data, then it's not that the statistics failed, but they were
misapplied, or accurate but irrelevant.

Bill Gates walks into a bar and a patron says to his buddy,
"Congratulations. We're all rich." His buddy asks him why, and he
responds, "The average person in this bar is now worth over a billion
dollars." His statement is true statistically, but irrelevant. In a case
like this, the relevant type of average would be the median and not the
mean. A statistician could figure that out easily, and it should be obvious
to everybody anyway. It's not that the statistics were wrong, but they were
used wrong.