"Joseph Gwinn" wrote in message
...

But I have to ask: Why is ice cream sales correlated with highway
accident rate? It is true; the question is why.

Do you mean by time of year? We drive more in the summertime,
especially
when it's hot enough for ice cream and to go to the beach.

Bingo! It's a very good example.

But see how easy it was to connect? Med-science researchers are far
better
than we laymen at recognizing coincidental correlations versus those that
have a strong chance of a causative relationship. Once they smell serious
smoke, the alarms should go off. Otherwise you're putting your life in
the
hands of people who just want you to buy their drugs and then go away.

Yes, this example is chosen to be tellingly obvious in retrospect, as a
teaching device. But respected researchers are forever getting tripped
up by unexpected common variables, and it isn't always so obvious. This
is one thing discussed at length in the NY Times Magazine article I
mentioned.

If this is what our discussion has come to -- the ideas that life if
complicated and that nobody is perfect -- I'd say we've stripped it all down
to something that no one could argue with. d8-)

--
Ed Huntress