The thing is, this poll does in fact meet the criteria for a push poll, not so much in the questions themselves, but the mere fact that only Obama voters were questioned. A theory about the knowledge of Obama voters, as well as a theory about the media bias, was formed, and the poll was designed from the ground up to gather data to support the theories. That doesn't mean the polling data was inaccurate, but is does mean the conclusions drawn from the data are highly suspect, specially if they are passed off as any kind of definitive conclusions, as John Ziegler is almost certain to do with his "documentary". Michael Moore made those kinds of leaps and bounds an art form in his documentaries.
The one question in the poll that was particularly striking is:
Which candidate said, "I can see Russia from my house"?
Considering that no candidate said that, and the only candidate who lives even remotely close to Russia is Palin, that question is a little like asking people when they stopped beating their wife. Stating that 25% would get that one correct by guessing is a misnomer, since 100% of the answers to that one would be incorrect, guessed or not.
The polling conclusions seem to hit hard on the fact that people know about Palin's pregnant teenage daughter, but then again she was the only candidate with a pregnant teenage daughter, and people knew about it. Ziegler is trying to make that out to be a media bias, but in order for that to be a media bias, Obama or Biden would have also had to have a pregnant teenage daughter that the press failed to report.
The biggest problem with the poll isn't that it's to gather data about the media bias, but it's to gather data to support the notion that Obama was elected solely as a result of the media bias. If the purpose of the poll was to actually reveal the media bias, then all voters would have been questioned, and not merely Obama voters.
Ziegler's defense of only questioning Obama voters is deeply flawed, ("I was not trying to prove that Obama voters are dumber or less informed than those who voted for McCain. I only polled Obama voters because I was trying to test the media's impact on the election. Since Obama won, it would be pointless (not to mention twice as expensive) to poll McCain voters.") as this only tests the media's impact on only one segment of voters.
The same questions asked of McCain voters would likely yield the same results in many cases. There's just an astounding percentage of people right across the board who are politically illiterate and get the bulk of their information from a slanted media. A more encompassing poll would have shown that, but then again, that wasn't the purpose of the poll, even though Ziegler has stated otherwise.
The studies have shown a particularly one-sided slant on this year's election. It used to be that while it was slanted, they at least made the cursory attempt at telling both sides. This year, they didn't even try. They went out of their way to slant the reporting and didn't try to hide it. Ziegler should have approached the polling for that reason, rather than concentrating on just the Obama voters. The end results would have been much more credible, just as accurate and, and much more resistant to the criticisms that it is getting.