On mathematical studies of the recall results part I

August 24, 2004

-In the comments below there are two simulations by ele and Gustavo, both of which claim that they get results of their simulations which agree with the results for the recall vote.


-I have received a Monte Carlo study by some Professors at UCV which reaches a different conclusion. Using data from 2135 centers, they use as the probability of participating the results of the CNE and the probability of voting “Si” the vote for each center based on the results of the 15th. They do 100 simulations using a binomial distribution and reach the following conclusion:


 


i)                    The probability of 236 centers having duplicate “Si” votes, as in the results, is less than 1%.


ii)                   The probability of having 6 centers having triplicate “Si” votes is less than 25%


iii)                 The probability of having one center having four machines with the same vote is less than 5%.


 


-A study of the results for the Actas per Center, when compared to the exit polls shows a low probability for the exit poll sample coming from the population of the actas.


 


More on all of these as I have more information or I am allowed to say more.

This entry was posted on August 24, 2004 at 11:43 pm and is filed under Uncategorized.

Leave a comment