|
The problem with the way these companies are performing the exit polls is that the error is unknown. If you take a completely random sample of all voters in a state the standard deviation is the square root of the number of people in the exit poll. This allows you to define the error.
The problem with exit polls is they are not a random sample and the error is unknown. They are not a random sample of the population because the variables include time and place. If you sample more in urban areas you get more Kerry voters, for example. If you sample during work hours you get a different voter than before or after work.
I am just guessing but the exit polls were probably conducted more in urban areas than rural ones.
__________________
skype = "adultdatelink"
|