|
Edited on Wed Dec-07-11 03:41 PM by karynnj
2008 - as various other teams challenged petitions that had problems. As it was not challenged, Indiana put him on the ballot and he and Hillary had a very close race with Hillary narrowly winning.
Here the numbers don't even make sense. I assume the point they are making is that the 150 was a sample of the 534. However, even if you assumed that the 2 that MAY have been forged actually were, you still end up with a result that it is unlikely that enough signatures would be eliminated. There was a 2/150% rate of them being bad, which would equate to an expected value of about 7 in the full 534. Assuming a binomial distribution, where p = 2/150, the standard deviation would be around 2.7, so the likelihood of there being 35 bad signatures is far outside any confidence interval normally computed. The upper end of the 95% confidence interval would be less than 15.
Note this assumes that the 2 are really fraudulent.
|