Discussion in 'Election Forums' started by Rinata, Sep 13, 2012.
Is it 2010? And how do you know the party affiliation of those who voted in 2010?
ROFLMAO......didn't bother to check their RIDICULOUS oversampling numbers did ya?
What do you think a proper sampling would be? And how'd you arrive at that figure?
Apparently every poll over samples democrats. I didn't think there were that many democrats left.
Ridiculous oversampling? The Florida numbers actually undersample Democrats (Democrats have a 4 point registration edge in the state, but only 2% more were interviewed in the poll). Not sure about the numbers for other states, but that one doesn't bode well for your thesis.
Maybe in order to figure out the appropriate sample we should do a random sample poll of a few thousand people in the state or something....
[ame=http://www.youtube.com/watch?v=_Rav9ijyyZk&feature=player_detailpage]Laugh track - YouTube[/ame]
Or look at voter registration data (which is what I did for the Florida numbers).
Romney has to stop the bleeding if he has any chance of winning. It's not past the point of no return yet, but it is getting close. If polling numbers move to Obama having a ten point lead or more, then it is over. Romney is desperate. The funny thing is that he keeps going off message and every time he does, it ends of costing him. His campaign team is a joke.
Either might give you a good answer but I'd argue a random sample poll is more accurate (Assuming it's done properly). In many states you don't need to declare a party for voter registration, which at any rate is typically not current data. But voter reg is far more accurate than "they oversampled because i feeeelllllll it can't be true!"
I'm just tired of hearing over and over again that polls "over sample" dems, while no one can seem to explain what the sample pool should be.
Separate names with a comma.