I think something like Presidential polling should be more straight forward that polling people about something they know nothing about. That being said, it takes no effort and there is not cost to lying to a pollster and even those that tell the truth it is unknown how many of them will actually get out and vote.
this territory is moderated
Yeah, and even with straightforward questions and luckily honest respondents, there's the whole issue of the sample of a thousand people actually being representative.
reply
Yes, we have to factor in the incentives that drive pollsters as well. If you are up against a deadline and you need a sampling of x y and z demographic but you are short some of z you may be inclined to assume some y's suddenly qualify as z's.
reply