via telephone - they never would have been able to contact me...
Yep! These polls are going to be more and more off if they keep using antiquated technology! ;-)
No seriously! landlines are pretty much gone. Cellphones made area codes irrelevant. Phone polls will be further and further off in the coming years!
Here are the problems with the poll:
1. The poll is registered voters, not likely voters.
2. The poll has a sample of 1012 registered voters, but the results are based upon a sub-sample of only 348 registered voters, and only 262 of that sub-sample identified as republicans.
3. The margin of error among subsets of the sub-sample is significantly higher than the overall margin of error of 5.3% for the sub-sample as a whole. (This should send up red flags.)
4. The 5.3% margin of error really means there is a 95% chance that the poll results are accurate within a 10.6% spread and a 5% chance that the poll results are completely screwed up (as opposed to merely suspect). In other words, the margin of error for this poll is 32.5% greater than a poll with a more typical 4% margin of error, and 76.7% greater than a poll with a 3% margin of error.