SPECIAL NOTE OF CAUTION #1: This poll does not and cannot reflect the views of all Americans. It only represents the views of people who watched the debate. SPECIAL NOTE OF CAUTION #2: The sample of debate-watchers in this poll were 31% Democratic and 33% Republican. That indicates that the sample of debate watchers is about eight points more Republican than an average CNN poll of all Americans, so the respondents were more Republican than the general public.
It’s in 2-point type, so yeah, easy to miss
So CNN itself claims that the results are not representative of the broader public, as determined by their concurrent national polling. This isn’t a matter of “unskewing” a sample to get numbers we’d prefer, but of CNN itself admitting that the partisan gap they’re seeing at the national level isn’t reflected here.
Of course, it could very well be true that more Republicans watched the debate than Democrats. That isn’t implausible, and would still mean the poll is accurate. So at that point, what would be helpful would be to see how Independents and undecided voters viewed the debate. And on that score? CNN blows it. Their poll memo does not include crosstabs.
Hence we have no broader context other than according to the Republican-heavy sample that watched the debate, those Republicans liked the Republican. If that’s what they were going to give us, they might as well have saved their money. I could’ve told them that for free.
Thus, the CBS News poll of undecided voters is more informative. CNN can rectify that by giving us better crosstabs.
Incidentally, that CBS poll shows the limited effect of a vice-presidential debate. While overall respondents said Biden won the debate by a 50-31 margin, nine percent were now more likely to support President Barack Obama (from 23 to 32 percent), while five percent were more likely to support Mitt Romney (from 28 to 33 percent). Not exactly game-changing.
On the other hand, Romney needs better than an even split to win this thing.