Proofiness - Charles Seife [47]
Journalists, at least the more sophisticated ones, are well aware of the huge problems with polling, yet they’re willing to ignore them. Even the dimmest of pollsters knows that Internet polls are utterly worthless. Ignoring the fact that Internet viewers aren’t representative of the general population, ignoring the volunteer bias caused by low participation in online polls, Internet polls are still so flawed that they have been considered the height of polling silliness for more than a decade. The major reason is that they can be easily manipulated with the slightest of effort, and often are. Internet polling is why “Hank, the Angry Drunken Dwarf” beat out Brad Pitt as People.com’s “Most Beautiful Person of the Year” in 1998. It’s also why Stephen Colbert, modern master of Internet survey manipulation, won polls to name a module of the International Space Station as well as a bridge in Hungary after him. Internet polls have no basis in reality whatsoever. Yet CNN.com has an Internet poll on its front page every day. It’s not for information; it’s for titillation.
Reporters are supposed to hold the truth sacred above all else, yet they don’t seem fazed at all when reporting polls that are utter tripe—so long as they have sufficiently small margins of error. They aren’t even bothered by the most glaring contradictions. At the very end of 2006, the Associated Press and AOL teamed up to conduct a poll designed to gauge Americans’ feelings about the coming year. As soon as the pollsters finished gathering the data, the Associated Press wrote not one but two stories about their poll.
The first: “AP Poll: Americans optimistic for 2007.”
The second: “Poll: Americans see doom, gloom for 2007.”
That’s right. Both stories were about exactly the same poll. Both stories were published by the same organization—the one that performed the poll. Both reporters consulted AP’s manager of news surveys, Trevor Tompson, before writing their articles. Yet the two stories used the same data to come to exactly the opposite conclusions. Nobody at the Associated Press seemed to worry that they were publishing nonsense; nobody seemed concerned in the slightest that reporters were presenting what I believe are meaningless, self-contradictory Potemkin numbers in the guise of objective truth.
In my opinion, there’s a legal term that describes this sort of behavior: absolute malice. This term makes journalists shudder, because it represents a failure so great that it pierces a hole in the First Amendment shield that protects the press. A reporter or news organization acts with absolute malice either when it publishes something that it knows is false, or when it acts with reckless disregard for the truth. When it comes to polls, I believe that the news media as a whole—not just the Associated Press—regularly act with reckless disregard for the truth.
Sure, news organizations try to dress up polls in the guise of objective reality; they publish a margin of error to tell the public how far the poll can be trusted. But at best, that margin of error leads to the sin of disestimation; since it doesn’t take into account systematic errors, it deliberately overstates the accuracy of the poll. At worst, it gives journalists license to ignore even the most obvious problems that plague a poll—it’s a tool they use to dress up nonsense in the guise of a fact. Journalists know that Internet polls are worthless, yet such polls regularly make the news. They insist on using telephone polls even though they know that such surveys are inherently biased.