A premature election orgasm...........via....AmericanThinker

"The key to scientific polls is random sampling of target groups, in this case (a) actual voters (b) Democrats vs. Republicans. If you’re not properly sampling those groups your poll results are going to be wrong....


Sampling bias would be serious if the pollsters missed even ten percent of the actual GOP voters. My unscientific guess is that they are missing much more than ten percent of conservatives, maybe 20 or 30 percent.

Any scientific paper based on such wildly unlikely sampling assumptions would be rejected by journal editors, and never appear in print. The media just don’t have anything approaching scientific standards. They are run by “activists,” not scientists. They’re already partying all over the international Left."