|
I think it was earlier in this thread that we discussed polling flaws and whatnot. The poitn was made about the pollsters being pros at their jobs and doing the best to adjust for the difficulties of random sample polling.
I agree with that idea, and the difficulty is that while you know some polls will be wrong, it is difficult to know which ones. Anyway, an interesting case in point.
Two surveys for Minnesota. Same polling period. Star Tribune has Obama up by 18 in that state. SurveyUSA has McCain up by 1. Anyway, found it interesting.
|