The News That's Not News

Each day, newspaper reporters and TV news pundits report the latest poll results -- suggesting that there has been a "change" since the day before. But the results being reported fall within the margin of error of the polling data: hence, the only real news is that headline writers are creating news out of wholecloth. For example, The Seattle Times "reported" today that "[d]espite polling that indicated Sen. John Kerry won the debates with President Bush, some surveys have shown the Republican incumbent gaining support."

The chart accompanying the paper version of the story shows data for Bush's approval ratings: 50% on 3 Oct; 47% on 10 Oct; 48% on 14 Oct; 51% on 18 Oct. Does this show "gaining support"? No, not when the margin of error is +/- 3%!

Reporting these data as Bush "gaining support" can create a bandwagon effect -- in other words, reporters and news editors can shape public opinion by presenting a false interpretation of the data. The same can be said for those who similarly interpret any Kerry "increase" that is within the margin of error.

Some foundation should buy a truckload of A Mathematician Reads the Newspaper and distribute copies free to newsrooms around the country. And journalism schools need to do a better job of educating future reporters about math and statistics. Bah, humbug.