Numbers Lie (Part 1?)
Fuzzy math pisses me off. Seems simple enough, but let me explain. There are any number of groups that are trying to convert us to their way of thinking. Some use rhetoric, some promises of gain, and some threats. All of these I can abide. What I can’t abide are the people who use phony numbers to try to persuade us. A few examples:Example 1: Seat belts save lives… probably true. Seat belts saved 11,000 lives last year… how the hell can you know that? A statement like that indicates that death can be predicted with certainty...it can’t. There is no way to know if the seatbelt saved someone or it was something else. A wreck happened her not long ago in which no one had a fastened seatbelt; one person died, the rest lived. If death were entirely dependent on the seatbelts, they should have all died. Dale Earnhart was wearing a five point racing harness, which is a lot more secure than a shoulder belt. Due to a combination of other factors, he died in what looked like a relatively minor crash. Without a full set of numbers, how many died in seatbelts, out of seatbelts, because of seatbelts (they do kill people,) etc., the 11,000 figure is just gibberish. It is an attempt to quantify a non event.
Example 2: Only 10 percent of rapes are reported (or the number I heard this morning on TV: only 1 in 40 rapes on college campuses are reported.) Again, how can you know that? While I do not doubt there is an underreporting of rapes both on and off campuses, how can you quantify a non event? And why would anyone be willing to make policy based on such ludicrous numbers? We do not know how many rapes go unreported for the precise reason that they are unreported. Any attempt to quantify the underreported amount is pure speculation. Most of these numbers rely on subjective, anecdotal evidence which is hardly the grounds for good statistics.
I will be posting more of these as they come up.
posted by Thomas @ 9:56 AM
0 comments
