The Democratic National Convention opens in Denver on Monday, followed a week later by the Republican National Convention in Minneapolis/St. Paul. So the last leg of a long and tumultuous presidential election begins. And with the final push toward Nov. 4 and a new president will come a multitude of polls -- so here I interject a cautionary note: Remember last winter, as the primary season began, when political pollsters couldn't get it right to save their lives -- or, in some cases, to save their jobs. Here's how media-watcher Howard Kurtz summed it up in the Washington Post the day after the New Hampshire primary: At the outset, the pundits seemed ticked that their expected story line -- an Obama blowout -- was failing to materialize. What about those pre-election polls we all based our blather on? When the cable networks couldn't predict at 8 p.m. that Hillary Clinton would lose, the commentators began wondering if she would declare herself the Comeback Kid -- as her husband did 16 years ago -- if she lost by "only" a few points. As the evening dragged on, the commentators had to consider the possibility that Hillary's "showing of vulnerability," as Tom Brokaw put it, might have helped her, and that Bill Clinton might have boosted her chances after all. In other words, that the coverage had missed the point. This was delicious. The coverage had been so out of control there was speculation about when Hillary might have to drop out. Polls giving Barack Obama an 8- or 10-point lead were accepted as fact. ... Some pundits were predicting a 20-point Obama margin. And then the voters actually went to the polls. The result: Dewey Defeats Truman. That may be overstating it a bit -- Clinton won New Hampshire with 39 percent of the primary vote, followed by Obama with 37 percent, John Edwards with 17 percent, Bill Richardson with 5 percent and Dennis Kucinich with 1 percent. But the point remains: We don't know for sure until the voters actually vote, despite what pollsters and pundits predict. The predictions, however, do make news. Steve Kircher, who recently retired after a long career as research manager at the St. Petersburg Times newspaper, pointed out in a recent Opinion piece that they should make news. After all, he notes that in the 26 Democratic primaries earlier this year, the polls were right in 23 of them. Only in the high-profile primaries in New Hampshire, California and Missouri were the pollsters dead wrong. Based on his own experience, Kircher offered what he called a "user's guide" for political polls. Here are the five elements he watches for: Low response rates: Virtually all national and regional surveys are done by phone, and the challenge is just reaching a person -- rather than an answering machine or voice mail -- who will agree to be interviewed. Since older people and women both answer the phone more and are more likely to agree to be interviewed, pollsters "weight" or adjust their data so this imbalance is corrected. Cell phones: Roughly 15 percent of American homes in 2007 had a cell phone but no landline phone, making the people who live there virtually invisible to telephone pollsters. Adding cell phones to a poll phone list is more complicated and costly -- and, in fact, can cost the person being surveyed if he or she pays for incoming calls. But according to Kircher, other surveys have shown that "cell-phone-only households are similar to landline households in political attitudes." Whether that will hold true as the number of cell-phone-only households grows is a question for a later poll. Lying: "As we can best determine," Kircher writes, "people are truthful in public opinion surveys, but there are times when the truth is stretched a bit. A recent Wall Street Journal story reported that in a 2005 Harris Poll 58 percent of interviewees told the live telephone interviewer that they exercise regularly, but in an online survey only 35 percent reported exercising regularly." In sensitive areas, like education, income and sex, people sometimes exaggerate -- but, Kircher notes, "That normally doesn't matter in political surveys." And race is definitely one of those sensitive areas. Kircher writes that some people have been "reluctant to say they're voting against a candidate because of race. As a result, polls tend to overstate the level of support for a black candidate by five to 10 points." Likely voters: "Nearly every national poll interviews only registered voters but some try to identify those people who are most likely to vote," Kircher writes. Those surveys "are considered to be more predictive of the actual election results because about six out of every 10 registered voters will go to the polls in November." The problem is that the likely-voter model tends to focus on older voters, and if turnout of younger voters is high, that could skew the results. Timing: Kircher blames "time-lag" for getting it so wrong in the three Democratic primaries earlier this year. The ones that turned out to be accurate included interviews up to the day before the election; the ones that were wrong included interviews from two to five days before the election. "If one day makes such a difference in polls results," he asks, "how accurate can polls be three months before an election?" Kircher cites one other important caveat: "I'm talking only about scientifically designed polls that use random sampling, not Internet surveys or other amateur efforts that do not purport to represent an entire electorate." * Sarah Jenkins is editor of the Yakima Herald-Republic. If you have a question or concern, you can reach her at 577-7703; P.O. Box 9668, Yakima WA 98909; or sjenkins@yakimaherald.com. You can also comment on this column in the "Inside the Newsroom" blog, at editor.yakimablogs.com.   The Democratic National Convention opens in Denver on Monday, followed a week later by the Republican National Convention in Minneapolis/St. Paul. So the last leg of a long and tumultuous presidential election begins. And with the final push toward Nov. 4 Valley Life, LOCAL

2008-01-09T18:15:55Z

by Sarah Jenkins

2008-08-24T02:42:51Z

false

2008-08-25T03:26:41Z

Election polls are getting harder to phone in 6999 election-polls-are-getting-harder-to-phone-in Web Story

JENKINS

Yakima Herald-Republic

2008-08-24T03:26:41Z

5966669

08/24/08 Sarah's Sunday column published Web Daily

2009-03-20T02:27:40Z