Misapplications Reviews: Newswatch

Everybody knows that some people lie with statistics and that others, while having no intention to deceive, nonetheless cite figures in misleading ways. But this knowledge has not stopped the widespread and often preposterous misuse of statistics in the newspapers and magazines and on the television and radio. And I am referring to data “analyses” that are transparently weak, not the unknown number of others that appear plausible but would collapse under scrutiny.

Let me not weaken the case by overstating it. Seldom do the data presented actually refute the point they were meant to support. Far more often, the problem is that numbers consistent with a wide range of hypotheses are offered as proof of the correctness of one of them. And with distressing frequency, the hypothesis “confirmed” in this manner is not the one the Goddess of Reason would have chosen in advance.