of guys in the office. The ones that did use numbers were often statistically dubious. The most common procedural error was something known as “selection bias,” a gaffe which has long plagued all kinds of polls, surveys, and scientific experiments. This occurs when researchers base their studies on samples that aren’t large enough, or random enough, to offer a representative cross section of the whole. The telltale sign was that most of these lists had a suspiciously regional flavor. Rankings from England, for example, were clogged with the names of soccer clubs like Liverpool and Manchester
...more
