Doug: Be very, very careful with those "Zogby Interactive" surveys. Unlike a traditional poll, which selects a sample by randomly dialing phone numbers, Zogby Interactive uses volunteers answering questions on the internet. While Zogby attempts to adjust the sample demographically to make it reflect a random, the self-selection problem is acute, and these surveys have not performed well at predicting elections.
In this specific case, the result does mirror Rasmussen's tracking poll of Obama's approval rating, which shows it hovering a little below 50%, though it departs from Gallup's tracking poll, which shows the President's ratings hover a little above 50. There's a methodological explanation for the discrepancy: Rasmussen filters its sample to capture only likely voters, while Gallup does not. Rasmussen's likely voter model has performed well in the past, so it would be fair to say that while Obama enjoys the approval of a slim majority of Americans, he does not enjoy the approval of a majority who are likely to vote (which is part of the reason why Democrats are likely to sustain quite a few losses in the 2010 elections).
Share this Article
Like this Article
Print this ArticlePrint Article