« Previous Post | Main | Next Post »
This poll, showing Obama pulling even with Hillary in Pennsylvania, cannot possibly be correct:
Pennsylvania DemocratsMar 7-8 / Mar 26-27 / Apr 5-6
Clinton 52% / 51% / 45%
Obama 41% / 39% / 45%
Someone else 1% / 2% / 4%
Undecided 6% / 8% / 6%
While I think the Bosnia lie did significant damage to Hillary's campaign, and Obama wasn't hurt nearly as bad as was initially suspected on the Radical Reverend (thanks in large part to the candidate's widely-hailed race speech), I think this much of a swing is beyond credibility and this poll is likely an outlier. That said, even a single-digit loss in Pennsylvania would be a win for Obama, and he appears to be building up an avalanche of support in North Carolina.
Posted by Kevin Hoffman at April 7, 2008 4:32 PM
« McCain shrugs off Secret Service: Bravery or campaign stunt? | Main | Alex Keaton, (Republican) America needs you now! »
As someone who has taken two years of graduate level statistics, I'd have to agree that it is probably an outlier. However, you'd have to know the polls alpha level which is not indicated on the website. A standard alpha is about .05 which means that there is a 95% chance that it's not just a freak number.
Posted by: Ben at April 9, 2008 3:38 PM