A poll done in April for the Mississippi Education Association showed how the state would have voted for Governor had the election had been held that day. The results broke down as follows:
35.4% Eaves
49.6% Barbour
15.0% Don't Know / No Reply
At this time nearly 1/3 of voters still had never heard of John Eaves while only 3% said that they didn't recognize Barbour's name.
This seems accurate enough, close to the numbers I would anticipate for the general election with the 15% not responding most likely not voting.
ReplyDeleteI have Fondren's one and only ad on tv on my page now.
ReplyDeletehttp://freerangepolitics.blogspot.com/
Eaves will pull more than that. He better!!!
ReplyDeleteWe'll see. I hope Eaves will do better than that!!
ReplyDeleteI would be surprised if Eaves kept the margin of defeat under 15 points.
ReplyDeleteNon-respondents do vote. They just don't want to commit on the telephone to strangers because some people fear reprisal.
ReplyDeleteActually, there are metrics that you can use in your polling to determine the people who are and aren't going to vote.
ReplyDeleteIf it's a serious poll, they'll have 2 numbers, one for registered and one for likely voters. Likely voters matter. If you're a likely voter and you're undecided, you're going to break one way or the other. An overwhelming majority (think like 99%, because otherwise, who'd pay 20k for a good poll?) of the sample completing the survey will give an accurate response. That is to say, if you're undecided, you say undecided. If you fear reprisal or have some concern about telling people who you're voting for because "it's supposed to be a secret", then you'll tell the pollster that and not take the survey.
This is only untrue in the case of final or late breaking polls. There's a dramatic margin, like 22% or something of 72 hour undecideds that actually vote even if they identify as likely voters. That is to say, as of your final polling on Friday night, you know how it's going to shape up. If you're within 3-5 either way, it's on your field program. If you're outside + or -, then you're made.
People, like my mother, like to doubt the validity and accuracy of polling. However, if it's a good poll, it's an accurate indication of the mind and tendencies of the sample. Like I said, if it didn't work, nobody'd pay 20k for a good one.
That's the word from your friendly neighborhood pol op.
I thought this was interesting especially with an election looming in the distance:
ReplyDelete"Govs. Haley Barbour of Mississippi and Mitch Daniels of Indiana have been subpoenaed by lawyers for people who blame their heart attacks or those of family members on the once-popular painkiller Vioxx.
Both governors have past ties to the drug industry, and plaintiffs' lawyer Russ Herman said they were subpoenaed to testify about consultations with the U.S. Food and Drug Administration before the agency set new drug label rules in 2006."
source: http://www.abcmoney.co.uk/news/262007108777.htm