Election Polls
|
2008 State Polls
|
Bill Richardson is Higher than Polls Claim Because of the Western BiasJuly 28, 2007 If your child got 90% or higher in every test throughout a school year but received a 45% average, would you not think there was something wrong? It could only happen was if there was another half of tests that you never saw where your child got 0%. Come on, how likely is that. I say this because this is what is happening with Bill Richardson's poll numbers. Half of all the state polls released in July shows him at 6% or higher but the national polls are still pegging him as this no-name lower tiered 2% candidate.
All of Richardson's national polling averages for July 2007 has been less than 4%. But if you take a look at the most recent state poll for those taken in either June or July, you would see that 11 of 16 polls show Bill Richardson is at or above 4%. The pollsters will claim that this all falls within the margin of error. The margins of error for the majority of polls is about 4% but that is the maximum error generally delegated to the winner of the poll, not the dude sitting at the bottom. The point is that there exists a strong likelihood that a candidate getting 32% is in the 28%-36% range but very unlikely that a 2% candidate is at 6%. Instead of talking up the statistical uncertainities in polls themselves and hiding behind the margin of error excuse, they should be looking at their polling methods.
I have a good deal of faith in state polls because a pollster can isolate and equalize certain important variables. A pure mathematician would say you could randomly select any 400 people and get a good polling measurement but that does not occur in practice because of the necessity to isolate variables like race and location. For example, voters in San Diego are more conservative than those voters in Los Angeles. Voters in San Francisco care are far more progressive than those in Riverside. San Diego voters care more about immigration than the rest of the state. Los Angeles has a crime and education problem that those in the rich counties near Marin know nothing about. By taking a poll of 300 people in California, you are able to spread those 300 voters evenly across the state and get a nice firm result. But if you do a national poll of 300 people, you would only be able to distribute about 37 of those to California (using the premise that California has 1/8 of the United States population). Wow. 37 people are determining the fate of California's representation in this national poll. The Western Bias. When a pollster is doing a national poll, they are distributing the voters they poll across the United States. But think about the time of the day that they call individuals. The perfect time to call an individual in the East coast would be working time for those in the West coast. The perfect time to call someone in the West coast would be sleeping time for those in the East coast. The majority of Americans are concentrated in those Eastern states so everything favors them. So in doing a national poll, they are forced to call the Western states closer to working hours when the base of supporters for Bill Richardson are likely to work. It is a fact that suburban voters are easier to reach than urban voters which is a reason why Hispanic voters in the West can tend to be disenfranchised, in the polls anyways. If you take a look at how Bill Richardson has been tracking in Western states, you can see that he is in a path forward.
USAEP Average. So what we did was assume that the state polls are the most accurate of all polls. And give each state a given number of delegates according to its population. Then we divvy up the delegates based on the percentage the candidate received in that state. Then finally find the proportion of the total number of delegates that candidate had received. Calculated based on percentage of total delegates.
Thus we project Bill Richardson is really at about 5% nationally and I am very certain that the national polls will start to reflect that. Who gets your vote in 2012?
|
Traffic During 2008 Election
2008 Election Results - Presidential
2008 Predicted Electoral Math
Latest Polls Per State
Poll of Polls Quinnipiac University Rasmussen Reports Polls Research 2000 Survey USA Polls CNN American Research Group Insider Advantage Zogby Mason Dixon Public Policy Polling Strategic Vision
2008 Democratic Primary Results
2008 Republican Primary Results
2008 Independent Candidates
Latest Presidential Tracking Polls 2008 |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Election Polls |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
