Since I do not follow pollsters and their methods that closely maybe someone can help me out. I've seen polls where they've surveyed say 1,000 Americans, 1,000 registered voters and 1,000 likely voters. I'm surmising that the same question asked to each of those three groups could have a different answer. Now I've heard of the big pollsters, Rasmussen, Gallup, Zogby, NYTimes/CBS, WSJ/NBC and there are probably a couple of others. Do they each survey the same group (meaning did they each only survey all Americans or only registered voters)? And if some survey only Americans and some survey only likely voters how are they differences accounted for when pollsters results are lined up next to each other? Thanks.