Note: written on Jan 15, 1996.

Humans have a saying, that one can lie with statistics, because numbers can be manipulated to support any argument. If one wants to demonstrate that the populace
is not starving, one adjusts the threshold where starvation sets in. If the numbers run up on one group don't look so good, pick another group. If the average is too
low or high, go for the median and arrange to discard the high or low end. Statistics, done honestly, can make a statement like no other, but done dishonestly are
deeply deceptive because the readership *believes* the numbers have been run up honestly.

In an era of increasing distress, governments want the statistics on the homeless, the unemployed, and the uninsured to *appear* healthy. Likewise, corporations
wishing to lie to consumers or to their stockholders discard the unpleasant from the computation and hope no one looks too closely. However, they are likewise
being increasingly challenged. What was included? How did you arrive at these figures? The squeeze is on. An easy out in these circumstances is to make the
formulas more complicated. Then the common man can't understand and the factors can be argued endlessly. One trick is to factor in a null, a zero, as a theoretical
possibility, when no such possibility in fact exists. Another trick is to hop through the data in intervals, taking a summation of spot testing, rather than a summation of
*all* the data. If hopping through the data with one interval doesn't give the desired results, try another interval. All in a day's work for the dishonest statistical analyst.