Regarding internals, the college does provide some, although not directly a party breakdown. At least, I didn’t see one.
Battleground Watch has posted that the poll is D+4. 2008 was D+6, 2004 was R+4. Thus, D+4 may be a little on the high side for the Ds. I think it will be closer to even. But D+4 isn’t indefensible.
So, we have a poll that seems to have reasonable internals, if a little biased to the left, that shows Gov. Romney winning clearly, and outside the margin of error. With 5% undecided. And among the 36% of undecideds who would say which way they were leaning, they were going Romney 3 - 1. But 4% identify a third-party candidate. It would be unlikely that the two or three third-party candidates receiving votes in this poll will actually amount to 4% of the vote on election day. When pressed to choose between the two major candidates, the third-party voters choose the Kenyan almost 3 - 2.
From all this, I’d guess Gov. Romney will take Virginia by about 10 points. Not going to be close.
One hesitation is that the sample isn’t terribly large.
sitetest
Thanks — I saw all the splits at the roanoke.edu site but they didn’t identify the partisan breakdown as far as I could tell.
If it really is D+4 and still shows Allen and Romney winning by as much as the poll refelcts, that’s fantastic but still enough of an outlier compared to other polls (even allegedly unbiased ones like Rasmussen) that a huge grain of salt must be taken.