Seriously, with any more than a few dozen respondents in the 18-24 category the chances of going from 55-40 Obama to 74-22 McCain are laughable. One in a billion? One in a trillion? Something like that.
Announcement
Collapse
No announcement yet.
Colin Powell endorses Barack Obama
Collapse
X
-
12-17-10 Mohamed Bouazizi NEVER FORGET
Stadtluft Macht Frei
Killing it is the new killing it
Ultima Ratio Regum
-
The drop I'm tal;king about is the 18-24 subcategory. Which, if it were in line with the other estimates I've seen would have put Obama back up at 7% or so (18-24 is about 10% of electorate IIRC, so a 60 point swing here means 6 points overall)Originally posted by Naked Gents Rut
The drop was only from Obama +6 to Obama +1 in a couple days.
To get a swing like that they either only called 18-24 year olds in Oklahoma or they had only 10 18-24s in their sample and got a number on the high side for McCain, then reweighted to the correct population.
The entire anomaly of that poll's headline number can be explained by a ridiculous anomaly in 18-24 year old category. Makes no sense. Is not statistics. Is bad poll-making.12-17-10 Mohamed Bouazizi NEVER FORGET
Stadtluft Macht Frei
Killing it is the new killing it
Ultima Ratio Regum
Comment
-
I wasn't using that poll to back up the point about American voters being fickle idiots swayed by the most trite and asinine talking points. That's an independent fact seen in countless elections. I just found the poll's progression in the past week separately interesting given their accuracy in 2004, but even that result was probably mere coincidence.Originally posted by KrazyHorse
Your real point should be that individual polls should not be trusted.
Comment
Comment