These polls are so fake. See, for example, this one that was just released, which shows ENTIRELY different results:
http://www.freerepublic.com/focus/news/3329442/posts
That’s why they average them. Like they say, there’s a margin of error. I’ll bet if you take the margin of error on each of them and add the margin to the 30% and subtract the margin from the 40% that they come close to overlapping.
Generally, the poll with the better weighting and the larger random sample will be closer to the actual views of the population their sampling.
This Gravis poll was a poll of registered voters...as you know such polls are generally more accurate than the Reuters or NBC or CNN polls that often just talk to “adults,” whatever than means.
I tried to look at the polling sample of the Reuters/Ipsos poll and couldn’t find anything about who or how many people they talked to....though maybe I just didn’t look in the right place.
Agreed. It’s way too early and besides, this current poll doesn’t even say what they pollsters asked. It could have been: who annoys you most? Who do you support? Who would you vote for against Hillary? We’re assuming the question but the article never says.