[Wolfstar]

I don't trust ANY one poll — not when they are up for a person or issue I favor, and not when they are down for the same person or issue. Polls are NOT scientific. To fool people into thinking they are, pollsters merely cloak themselves in the legitimate mathematical discipline of statistical analysis. However, it is extraordiniarly easy to skew poll results with:

->question wording,

->question order,

->day of the week and time of day a respondent is questioned,

->who the respondents are (adults, registered voters, likely voters, etc.),

->the number of questions asked,

->sample size,

->weighting required to "ensure" randomness due to high number of refusals to participate,

->weighting required to "ensure" "fair" representation of different political points of view.

At best, a group of polls taken over a period of time and averaged, might give an approximation of genuine public attitudes on the subject polled.

[irish guard]

>>>>>>.At best, a group of polls taken over a period of time and averaged, might give an approximation of genuine public attitudes on the subject polled<<<<<<<

If that werew true, then Bush would have won in 2000 by about 5% and would have taken a number of states that went for Gore. I agree about the wording question order and all that, but Zogby is good. The sad truth is Zogby was right in 2000, Bush coasted on his lead and almost lost.