It's really amazing how a thread that started by Meatball on his
quitting of smoking -- turned into a tree-hugging crusade by Greg
Mossman, who has absolutely no knowledge about how CAUSATION can
be ascertained statistically, exhibited his gullable in any and all
anti-smoking propaganda.

Greg was glad to change the subject to LAW (which he thought he knew
something about) when he was alone on a limb on smoking, without any
leg to stand on, and was rudely awaken that he didn't even know the
most basic facts about LAW -- the court standards!

Once he hastily took the bet that he was correct in his earlier
(May 27, 2004) argument, that "beyond a reasonable doubt" was NOT
as high a standard as the lesser standard of "clear, convincing, and
unequivocal" evidence, he IMMEDIATELY lost the $1,000 bet he accepted,
and started weaseling his way ever since.

Greg made a SECOND blunder, by falsely claiming there are "three
and only THREE" standards, while it was clearly established in a
referenced paper on FOUR standards and data on 10 judges on those
FOUR standards.

Then Greg went on a tangent to bet on his verifiable record of an
IQ test score, when he has NONE to show!  He had been polluting
(pun very much intended) that smoking thread by playing KING Weasel
ever since.

I am damned tired of Greg's weaseling acts, and am ready to get
BACK to the "causation" issue about smoking.

We have a lawyer who knows NOTHING about statistics arguing about
causal inference in statistics, claiming that a controlled experiment
is NOT necessary.  Correlation and common sense would suffice, says
Greg.

Little did Greg know that his argument had already been shot down by
me, OVER 30 years ago, in a 1972 publication:

RF> For a serious discussion of this aspect, you may read what I've
RF> published in a 1972 issue of the Journal of the American
Association,
RF> the review of a book titled "Correlation and Causation", in which
RF> the author presented his case that "correlation" studies (such as
RF> your figures on rates and associations) can establish/prove
RF> causation. I gave my statistical argument why, in the absence of
RF> a controlled experiment, such a conclusion is unwarranted, and
RF> the correlation methodology presented in the book was just "black
RF> magic" (I used those words). Don't recall if I actually used the
RF> word "quack" or not.

RF> That's about it.   Every other week you read some news item like
RF> "Study links escargo eating to faster running", or some such, and
RF> the only thing it says is that someone got some grant from the
RF> government, and found some spurious association and called those
RF> "links".

The pool is open.

-- Bob.