Tuesday, October 14, 2008

Polls

So we have all heard that Obama has pulled ahead by 9 points in the latest polls. As I was reading this article about it, a few questions came to mind. I'm not sure who might be able to answer some of these, but I'll throw them out anyway.
  • This particular poll interviewed just over 1500 people, not all of whom were even registered and even less who were "deemed likely to vote." Clearly the idea is that these people are random and somehow representative of the population as a whole. I am curious to the logistics of this. How are these people chosen and contacted? I would just like to know more about the process.
  • In past elections how accurate have polls at this point (3 weeks out) been in predicting the actual outcome?
  • Have their been times when one candidate had a substantial (or comparable to Obama) lead 3 weeks out and lost?

I am going to see if I can find some answers to those questions so I'll keep you updated.

1 comment:

joeydee said...

Hmmm, yeah. I'm wondering about a poll in which a third of the participants aren't likely to vote. Not sure if that's unusual or not, but I would be like, who cares about your opinion if you're not going to vote? You could say anything.