There was once an article in the Atlantic Monthly (I think it was then
called the Atlantic Monthly) entitled "There are XXX trees in Russia".

The writer dealt with the habit of journalists leaving blanks in their
copy, which researchers were later supposed to fill in. The point
being: how meaningful could a statistic be, if the writer didn't know
it to begin with?

The writer's favorite case involved the trees-in-Russia question.
Let's say it was Newsweek. A Newsweek researcher naturally called the
Russian embassy, which admitted it didn't have a clue. The researcher
then estimated the acres of forest in Russia, estimated how many trees
per acre, and multiplied the two figures. Newsweek published the
number.

The next week, the Russian embassy called up, pleased as punch, and
announced that it had the answer to the researcher's question: there
were exactly XXX trees in Russia.

(The source, of course, was the Newsweek article, a perfect example of
the referential nature of the "facts" we get in the news.)

all the best -- Dan Ford
email:

see the Warbird's Forum at www.warbirdforum.com
and the Piper Cub Forum at www.pipercubforum.com