“The survey of 925 registered voters had a margin of error of 3%; the Republican sub-sample of 363 voters had a margin of error of 5%. It was conducted jointly by Democratic pollster Anderson Robbins Research and Republican pollster Shaw & Co. Research.” Only 39% of the sample were republicans. Were the other 61% democrats, making this a heavily left biased poll? Was it split 39% republican, 10% independent, 51% democrats, still making it much more left biased than the national average? The combinations are endless, and the sample size so relatively small as to be almost meaningless. And yet these numbers are quoted as fact and representative of the nation.
No comments:
Post a Comment