|
According to the publicly available methodology statement for the 2004 exit polls, only 250 polling places were polled. This is about 2% of the total of 11,719 polling places. In other words, they polled an average of 5 polling places in each state. For reasons I spell out below, it is important to find out where these places were, and it should not be difficult to find out. With this information, it may be possible to show that fraud has been committed. I have only recently been given permission to open a thread here, and I have been perplexed and frustrated at the lack of attempts (until recently) to discover this important information, and the reliance on incomplete data and unfounded claims.
It is not stated whether larger states had more areas polled than smaller states. I would expect so, but not proportionately. These places were chosen as a 'stratified probability sample' of each state. In other words they were not chosen randomly, and therefore the pollees could not be said to be part of a true random sample. Apart from only 500 absentee and early voters contacted, by phone, in only 13 states, anyone not in those 250 polling places were ignored. There has been some discussion on this list as to the actual numbers of voters polled, but at this point, I don't think that is relevant: what we need to know is where these polls took place.
It is not publicly known which 250 polling places were exit-polled. It is obvious, however, that if we want to measure the accuracy of these exit polls, we have to discover which places these were, and then compare the exit poll results with the actual results IN THOSE PLACES WHICH WERE POLLED.
It really should be very easy to discover what places were polled. Just ask around if anybody was polled in their area. I'm sure that some agencies have already done that. And, of course, I'm sure that Mitofsky et al have given the information to some people. The information may well be available somewhere, but I have not been able to find it. It may even be tucked away somewhere on DU, I suppose. If indeed the polled areas are known, I would apologize for my ignorance, and beg enlightenment.
The result data that has been provided, by CNN or 'Simon', or anybody else does not mention which areas being polled, and there is no attempt to break them down. Any analysis based on these figures, comparing them with the nationwide results, is incomplete and misleading. As this flaw is obvious even to the most elementary student of statistics, I would even go as far as to say that in some cases these analyses have been deliberate misinformation.
Prior to the election, the pollsters would have kept the location of the polling places a strict secret, so as to avoid manipuilation of the polls. I'm sure that any party trying to manipulate the computer counting of the votes would have tried to discover this secret, and therefore keep away from areas where the manipulations could be discovered. I would not be surprised if anybody in a postion to manipulate the vote would also be in a position to discover the polling secrets. However, it is possible that they were not, in which case the smoking gun might still be there in the exit polls.
The fact that so many exit polls showed an advantage to Kerry is NOT a smoking gun at all. It is not even suspicious, as there is every possibility that the stratified probability sample of polling places produced proportionately more Kerry-leaning areas. (Smaller, rural places are less likely to be chosen). As long as the statistics of each area is well known, this would not matter. What DOES matter is whether the polling data corresponds to the actual results in the areas polled. As the voters were chosen at random, one would expect the results to be very close to the actual: even within point five percent. As there would be no reason to suspect the exit polling at this stage, any large deviation would have to point to manipulation of the vote, and if that deviation is repeated in multiple areas, then we can, of course, be very suspicious indeed.
There are a couple of things which really disconcert me. The first is the bullshit that Mitofsky came out with when trying to explain the apparent contradictions in the polling. By saying that, for example, too many women were polled, he is essentially saying that it was a very badly run exit poll, as that is just something that should not have been allowed to happen. If this were the case, and I can't believe that it was, then his employers should ask for their money back.
The second thing is that they have taken so long to come out with fuller information regarding the areas polled. It is normal commercial practice to keep these things under wraps until a full report is produced, but I would have thought that if there were nothing to hide, and every incentive to put an end to rumours, then they would have produced the data. While I'm fairly sure that if the data were damning it would have already leaked by now, and the Democratic side would have exploited it, I still have my doubts.
In ending, I should say that I am not in a position to find out these things myself. I am not in the USA, and not American. I'm British. However, I'm sure you understand that the results of the USA election are important to any nationality.
I write here because I feel concerned, but I just do not have time to take part in the kind of discussion and QA that often takes place on this list. Please don't think me rude. I just don't have the time to respond, and have done my best to write what I think in one big message. If you believe that I am in error, or am incomplete, that is fine. I'm always happy for enlightenment. Just please do not try to tell me that you already have the 'smoking gun', because you haven't.
|
Printer-friendly format
Email this thread to a friend
Bookmark this thread
(17 posts)
Send PM |
Profile |
Ignore
