Never Take A Poll At Face Value

Polls are tools if you understand them correctly they can give a wealth of insight into an issue. But never take them at face value alone, now what the margin of error means, what the actual questions and answer choices That were used for the poll.
For example, I was once asked to participate in an over the phone poll. I was asked how I thought a certain local politician was doing his job. But rather than being able to say I didn't think he was doing a good job at all, I was only given a choice between a) a good job , or B) a fair job. There was no C) for just plain bad.
Poll questions can also be worded so that they are certain to get the response that the poll taker wants.
Why do they do this? So they can wave a poll in front of you hoping to sway your opinion. Any one in advertising , marketing, and politics knows this, and use it. Some are legit, some are bias, some are just poorly made. You need to know the difference and what they are really telling you.

clipped from rhetorica.net But there’s a bigger problem: the reporting of polling by journalists and bloggers. All too often the reporting fails to accurately describe the margins of error or the questions asked. Without these adequately explained, any report of a poll is pure (and dangerous) nonsense. In a poll with a typical margin of error of 3.5 points, anything within the margin is a tie. Now you have to know what question was asked and how it was asked and to whom it was asked. If you don’t know these things, then you still don’t know what the poll means. But the poll will have meaning. The media (mainstream and otherwise) will give it meaning. And if people act in political ways based on the meaning given, then that’s what the poll means. Get my meaning? Why do the polls disagree so much? And which ones should you buy? It seems to me that journalists should not report on polls unless they understand and present the hidden ingredients.