Just noticed that this is RV. Why is anyone doing RV at this stage?
...no?
"A total of 820 registered Florida voters were interviewed statewide by telephone. All indicated they were
likely to vote in the November general election."
A voter's indication that he/she is likely to vote isn't a reliable LV screen.
!
I'm fine with it if that's what MD wants to use, I just prefer stricter LV screens. It stands to reason that if you call 800+ people, chances are that not every one is going to vote.
But admittedly, I missed that sentence in the article. It's probably a similar LV screen to what some other pollsters use (in fact, PPP used the same screen in 2008, I believe, when they got stellar results). In a super high turnout situation it works well.
