CJ,
I have another solution. Voting could be like jury duty. x number of people are chosen completely at random to vote. This would fairly sample the population.
In a two person election if you sample 10,000 people the error is 1%. This means that polling 10,000 would correctly sample the population to within 2% in 19 out of 20 elections.*
* The standard deviation of 10,000 is the square root of 10,000 which is 100. This is 1%.
So in this case two standard deviations is 2%. Two standard deviations encompasses 95% of the population (19 out of 20).
|