“Now, before you go all Rush Limbaugh and start talking about how all the pollsters are biased, consider that Gallup had Bush ahead by far more than his actual winning margin in 2004.”
I wouldn’t say all pollsters are biased, just ALL the polls. The pollsters even brag about it, you just don’t understand what they are saying. First, let me point out that Obama being ahead 51 to 47 doesn’t mean he will win, it means that if they held 100 elections, Ohhhhhh……BAMA! would win 51, NcCain would win 47 and 2 would end up no results.
Polls give ODDS not predictions. The Pollsters use various algorithms to “spin” the results. They are trying to produce a prediction. Statistical Analysis is a scientific tool used by engineers and scientists in various disciplines. It is the father of Opinion polling. Opinion polling differs from statistical analysis in that Opinion polling is about people, while stats is about physics. People LIE and change their minds. Physics doesn’t lie and always stays the same or changes by established rules.
The Law of gravity won’t change tomorrow because somebody spills coffee in their lap on the way to work. An opinion might.
Here is one guys opinion of how Gallop “spins” their polls;
http://www.thedemocraticstrategist.org/donkeyrising/2004/10/gallup_poll_raciallybiased.html
This is worth snipping;
“Note that the demographics of Gallup’s LV sample are not adjusted in any way (as their overall samples are) and are simply allowed to fall where they may.
What this means is that if, say, minority voters are much less likely to answer the 7 questions “right”, they will be correspondingly under-represented in the LV sample–perhaps severely under-represented.”
This applies across ALL demographics. In a contest like this one, where Race has been injected as a basic issue by the candidate that is a minority, Demographics will be critical. Yet Gallop is adjusting ( spinning) for political party. We will see how that works out on November 04.
