|One question point blank asked if I supported or opposed "the right to work." No explanation. A lot of average voters won't know what that means, when asked off the cuff like that, and of course they'll say "yes" without understanding the question.
Before that, a series of questions asked me to rate on a scale of zero to 100, with zero meaning I'd be least likely to support the candidate described, and 100 the most likely, my thoughts of a candidate with several specfied descriptions.
A few of these questions were largely unbiased, like "the candidate is strongly pro-life." Everyone understands abortion politics well enough to answer appropriately according to their personal views on the issue.
But other questions were very leading and hard to answer in the negative, by their very nature. One focused on "the candidate" did such and such to change state ethics laws to do politicking while he was a State Senator. McDonnell obviously is testing whether he can manufacture a scandal out of something most people aren't even aware of, including a political junkie like me. The candidate "description" was phrased in such a strongly critical way that even I couldn't say "100," meaning I'd be sure to support a candidate with that description. So I answered "50," meaning no more or less likely to vote for him...which is true, given I know they're talking about Deeds and I'll simply ignore any manufactured "scandals." But I'm a "strong Democrat" and "liberal" as I answered in a couple other questions, and people of other political persuasions are going to answer "zero."
The problem with some of these message-testing questions, like the one I just tried to describe, is that they essentially ask "Would you be more likely or less likely to vote for a candidate who beats his wife?" What the hell will anyone say?! And yet it doesn't really test a message in the dynamic of a real campaign, where there is pushback. I know campaign professionals must see value in questions like these, but I think they're mistaken.
I contrast this to some other more neutral message-testing questions the poll asked (without the zero to 100 scale), such as on spending cuts versus tax hikes to balance the budget--some of these policy questions phrased both alternatives in a largely positive light and thus were unbiased. These questions reflected how messages get tested in a real campaign, with pushback and counter-arguments.
A final note: the live caller asked for me by name up front, and at the end of the poll again confirmed I am the named respondent they sought. That's also when, per Virginia law, they revealed the poll is paid for by the RNC, which I was going to ask if it wasn't volunteered.
I wish I had the opportunity to take notes, but I was distracted by my 3-year old daughter nipping at my heels and trying to get my attention. No way, caught off guard as I was about to put her to bed with my wife out for the evening, could I get pen and paper and jot down the most interesting questions.
I don't know if there's any value in trying to get some of this information to the Deeds campaign, especially since I wasn't able to take notes, but maybe I'll try.