- Sep 25, 2011
- Reaction score
- Fredericksburg, VA
So you are saying that a 1000 sample poll in Rhode Island is no more valid than a same size sample for the whole worlds population? Seriously?That's not how the math works.
Sampling is not based on a percentage for a large population. It's based on an absolute amount that is larger so that the expected variation around a mean declines. This can only happen with a sizeable sampling.
That is nonsense. You need to have a demographically matched sample to the population you are polling, and to do that you have to have a significant sample from each demographic.
Do you think these national polls were not cross indexed with demographic stats?
No, most of them were not, and THAT is why they are so inaccurate. The USA is not 48% Democrat vrs 37% Republican, so their results are way off.