"Spehro Pefhany" wrote in message
...

That's the problem with statistical analysis as opposed to using the
scientifically sound experiments with control groups etc. It is a heck
of a lot cheaper and easier to use statistics, but you can only show
correlation. Correlation does not prove causality.

You should hang around rec.boats a few times a week and listen to the
******s who say "Well...since we invaded Iraq, we haven't had any terrorist
attacks here, so the war must be working." It's enough to make you puke.