Badger
badger
Atlas Legend
Posts: 40,501
|
|
« on: October 09, 2012, 08:32:01 AM » |
|
I'm looking at the "raw data" from the internals, and it appears those WERE closely approximate the numbers used for their top line results. e.g. Romneys 51-43 lead among men and tied at 47 among woman average out to the 49-45 lead reported. The other subsample "raw numbers" also tend to correlate these calculations.
So this poll's results indeed appear based on Pew's sample being 83% non-Hispanic white (and only 5% Hispanic!) and having more seniors than all voters under 50 (including only 7.5% under 30). Not at all an an accurate model of what the composite vote will actually be, to put it mildly.
If that encourages Obama supporters to simply whistle past the graveyard about this poll, two other things should be noted. First, the weighting towards votes over 50 somewhat HELPS Obama in this poll as he leads handily among 50-64 year olds (36% of the total sample). Secondly, even if this poll is an outlier, there's obviously a national shift to Romney reflected in most other polls. It's uncertain whether the good jobs report has helped blunt some much needed enthusiasm for Romney among the base, and even among independents now willing to give him a second look.
|