Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Politics : President Barack Obama -- Ignore unavailable to you. Want to Upgrade?

To: nigel bates who wrote (44158)	11/10/2008 7:53:40 AM
From: zeta1961	1 Recommendation Read Replies (2) \| Respond to of 149317

Here's Ann Selzer on her polling.. My view Worry drives pollster to get Indiana right J. Ann Selzer Posted: November 9, 2008 Of the three states I polled in this general election, I worried most about Indiana. My polling firm serves three major newspapers: The Des Moines Register, The Detroit Free Press and The Indianapolis Star. Iowa went for President Bush by a small and late-breaking margin in 2004 and so it was considered a battleground state in this election. Our polls never confirmed that status. In February, after John McCain sealed the Republican nomination, we polled likely November voters on an Obama/McCain matchup. We published the same 17-point margin of victory for Barack Obama then that we published Sunday in our final poll. Iowa never seemed in play with Obama as the nominee; McCain had little chance of holding this state that was red for only a short time. Obama took Iowa by nine percentage points. Advertisement Michigan was also thought to be a state McCain could win. Looking at our polling data, I was never sure why. And that was confirmed with a wide victory for Obama Tuesday (57 percent to 41 percent, the same 16-point margin we predicted in our final poll published Sunday). So, I polled in two states thought to be battlegrounds that were not. Then there is Indiana. A few months ago, no one thought Indiana could be in play. Hoosiers gave George Bush a 21-point margin in 2004. Our polling in September, however, confirmed that indeed, Indiana was very much in play. The secret to my success as a pollster is that I know how much to worry and where to direct that worry-driven energy. I've been worrying about Indiana for months. We predicted a tight race with Obama one point up, and that is exactly what happened. We predicted Gov. Mitch Daniels would sail to an easy 18-point victory; the final margin was 18 points. A few other findings from our final poll panned out, and our findings by race are perhaps the most important to explain. We estimated the black turnout at 8 percent of the total vote. It turned out to be 7 percent, the same percentage as 2004. We also showed 82 percent of the black vote going for Obama with 10 percent undecided; it turned out to be 90 percent. We knew independents would be a critical voter bloc as well. We showed Obama leading by 11 points; that's exactly what happened Tuesday. How were we able to get this right? We did not, as some pollsters do, look at the last general election and trust it to reflect the current state of the electorate. We do not want to get in the way of seeing real change when it is there. We knew the African-American vote would be crucial to getting this projection right. We also knew there was intense interest and unprecedented organization in the black community. It would have been easy to overestimate the black vote. That would be just as much a defect in our poll as underestimating the vote. We do not make adjustments by party, though that is done by many pollsters. To me, party affinity is a moving target with no firm benchmark against which to judge. So, I keep my hands off our data when it comes to party preference. We did take extra steps in Indiana to get this final poll right. We asked for demographic information from everyone we interviewed -- whether a likely voter or not. That way, we could compare the profile of our full respondent pool to the profile of the voting age population provided by the U.S. Census -- definitely a firm benchmark. This enabled us to make adjustments without making guesses about what groups would be more or less likely to vote. When we pulled out likely voters from this larger sample, groups that were more likely to vote would be properly represented, along with groups less likely to vote. That kept us from having to make gut-level decisions about the electorate. We did one more thing in Indiana. We identified a group of what I'll call "somewhat likely voters." Normally, we only interview individuals who tell us they will definitely vote on Election Day or have already voted by absentee ballot or early voting. In this last poll, we kept the small group of somewhat likely voters on the phone long enough to find out whom they would support. In reality, we were looking for any late-breaking McCain vote, since all the national polls defined Obama supporters as highly enthusiastic and motivated. And, we thought we might find something telling about how race affected voters' choices. What we found in our poll, however, was that this group, if anything, broke for Obama. The exit poll further clarifies that perceptions about race were not pivotal in this election. More than one in four voters (29 percent) say the race of the candidate was a factor, with 70 percent saying it was not a factor. These two groups of voters voted identically: 50 percent for Obama and 49 percent for McCain. And that is a bit about how we got it right. We relied on data-driven methods for identifying an accurate cross-section of who turned out on Election Day. Polling gets harder year by year. I have to end these comments by thanking all who agreed to participate in this tool that helps us understand our democracy by answering your phone and agreeing to be interviewed. Without you, we'd never stand a chance. • Selzer is president of Selzer & Company, the polling firm hired by The Star and WTHR (Channel 13).