Interesting article over on Slate by William Saletan on polling methods. You can read the whole thing by clicking here or I can just summarize it for you: The results of a public opinion poll can vary widely depending on precisely how the question is phrased. That's really all the piece says but the examples are rather interesting.