Thursday, May 26, 2011

Welcome to the Bad Statistis Blog

A short introduction: Some years ago (well, in the middle of the last century, actually), I spotted a clearly inaccurate numerical reference in a news article. The article was written by the Science Editor (I don't remember his name) of the Los Angeles Times - at the time an newspaper of pretty-good-repute. The article was about geothermal energy, a hot new topic at the time (heh). In the course of the article the writer quoted a U.K. geophysisist (I don't recall his name, either) as saying to him something like this: "We've found useful amounts of geothermal energy in test wells as shallow as 8000 Kilometers".  
Well! I was a bit gobsmacked at this claim. It's wrong in so many ways it's hard to know where to begin!
The simplest explanation is that the writer simply didn't bother with such numeric trivia as the difference between a meter and a kilometer. The difference, as it were, between one and one thousand
In succeeding years I've noticed that this is far from an unusual thing. I often see numeric references that are off  by somewhere between one and several orders of magnitude. It seems clear (to me, anyway) that if the numbers are wrong, then the conclusions drawn from them can hardly be correct.
This blog will try to collect some of the more interesting errors. A complete compendium of the misstatements, misrecordings, and out and out mischaracterizations of numeric data is, of course, impossible in these times, when so many seem so devoted to stating wishes as if they were facts.
I will try within my own limited abilities to be fact-based.

Thank you,
Bad Statistician (26 May, 2011)