Written by
An old children s math/logic joke goes something like this:
Three men on a business trip decide to split the cost of a single hotel room to save money. The desk clerk tells them the cost of the room is twenty five dollars. Each man gives the clerk a ten dollar bill, equaling, of course, thirty dollars.
The clerk does not have correct change, so he advises the three men to go ahead to their room, and that he will send the bellboy up directly with the five dollars.
So, as promised, the clerk gives the bellboy five one dollar bills, with instructions to divvy it up between the three men. On his way up to the men s room, he can't figure out how to divide five ones between three men, so he decides to give each man one dollar, and pocket the other two, thinking they wouldn't notice.
Each man had paid with a ten, and got one back, making nine dollars per man. Nine dollars times three equals twenty seven dollars, plus the two the bellboy kept makes twenty nine dollars. Where did the other dollar go?
The point to my telling this, is, don't be fooled by polls, statistics, and studies. Numbers can be made to represent almost anything the statistician wants them to represent.
A recent case was trying to prove that abstinence does not work because protestant people have a higher number of unwanted pregnancy, teen pregnancy, and abortion than non-religious women.
All of the studies cited in the arguments seemed to make a pretty good case, until you looked at it from a logical viewpoint. They compare the numbers on a per capita rate, or break the numbers down by the number per religion, not taking into account that the religious people are the vast majority of the population, and therefore, all other things being equal, should have the vast majority of abortions, teen pregnancies, etc., which, of course, they don't.
For any study you see, you can probably find another which contradicts it, simply by attacking it from another angle, or skewing the information by omission of the actual measuring stick.
A novice's guide to understanding the ways statistics can mislead is linked on the left.
The book, "How to Lie With Statistics", by Darrell Huff, was published in 1954. It has sold over a million and and a half copies, and deals with subjects such as "corelation does not imply causation" and "using random sampling".
Mark Twain borrowed a phrase to express his feelings about the matter.
"Figures often beguile me," he wrote, "particularly when I have the arranging of them myself; in which case the remark attributed to Disraeli would often apply with justice and force: 'There are three kinds of lies: lies, damned lies, and statistics.'"