You're quite right about that, bb. I am not saying "definitive" in the sense that it MUST be right. I was just pointing out that I HAD given the exact source of my data.

That said, if you read what Zogby has to say about the breakdown of who was polled, you will find that it is not done that way to gerrymander the results. Zogby uses the breakdown he does (38 % registered Democrat, 33 % registered Republican, and the rest independent) because that's the breakdown of people who voted in the last election. Other organizations change this around, in theory because they believe from previous polls that the breakdown of voters in the coming election will be different. Are they trying to play with the results? It's possible.

If you think about it, though, the independent pollsters like Zogby really have no vested interest in misreporting. Sure, there is a vague possibility that they could themselves affect the outcome of an election. But if you weigh against that all the years it takes them to build up a reputation, it's a mighty big risk. They'll start trying to affect the election rather than accurately predict it, their accuracy will go down, and pretty soon their reputation will go with it. I think that's why Zogby simply uses the breakdown of voters in the last election: it's the most transparent method.

Other polls being run by, say, CNN, Time magazine, etc. MAY be more suspect. After all, they did not build up a reputation primarily as pollsters, and their reputation is less dependent on their performance in that area.

Oh, and there's this: of all the current pollsters, Zogby has most accurately predicted the outcome of recent elections.

Just something to think about....