I was looking at the polls from the past couple of days, and I have a question:
My question is, how can they claim these to be statistically accurate to those margins when they differ by so much? The CBS News results are totally outside the margin of error on the other two posts. Is the problem statistically flawed analysis, or are one (or more) of them just making things up?
- CBS News: Obama up by 14% (moe +/- 3%)
- Zogby: Obama up by 3.8% (moe +/- 2.9%%
- Rasmussen: Obama up by 5% (moe +/- 2%)
My question is, how can they claim these to be statistically accurate to those margins when they differ by so much? The CBS News results are totally outside the margin of error on the other two posts. Is the problem statistically flawed analysis, or are one (or more) of them just making things up?
- Mood:
confused


Comments
I expect that they also might not have samples that are true cross-sections of the populations...
* Zogby: Obama up by 3.8% (moe +/- 2.9%% with single parents on welfare.
* Rasmussen: Obama up by 5% (moe +/- 2%) with atheists who love Joe Biden's "'tude".