Missing In Action?

[crimresearch]

Anyone else notice that after all of the name calling by KevinLowe and Ion against anyone who merely asked for a detailed analysis of the Ohio voting issues, they are nowhere to be found when this comes out?
--------------------------------------------------------------------------------
Analysis of the 2004 Presidential Election
Exit Poll Discrepancies
"There is already a strong case that there were significant irregularities in the presidential vote count from the 2004 election.
Nevertheless, critics are asking for firmer proof before going
forward with a thorough investigation30. We feel strongly that this is the wrong standard.
One cannot have proof before an investigation.
In fact, the burden of proof should be to show that the election process is accurate and fair.
The integrity of the American electoral system can and should be beyond reproach. Citizens in the world’s oldest and greatest democracy should be provided every assurance that the
mechanisms they have put in place to count our votes are fair and accurate. The legitimacy of our elected leaders depends upon it."

http://electionarchive.org/ucvAnalysis/US/Exit_Polls_2004_Edison-Mitofsky.pdf
---------------------------------------------------------------------------------

I guess some people were only here to stir up conspiracy theories and troll, not to examine any facts.

 

[Beerina]

The people who study these things leave a lot to be desired in the brain department.

For example, they only have a small, finite number of exit pollsters. Since each pollster takes x minutes per poll, they'll undersample heavy voting periods and oversample (relative to heavy) the light voting periods. To correct for this, they would have to have someone counting all people exiting, then do a weighted scaling of the polling at that moment vs. the number of people exiting in that period of time.

To put it more brutally, if, for whatever reason, Republican voters (being more likely to be employed, say, for the sake of argument) voted early morning or in the evening, then polling taken at those points in time might represent 50% of total voting. Thus the exit poll ratios at that point in time should be scaled to 50%.

Say they took 20 polls during the morning and 20 during the evening for a precinct, and took 60 during the middle light day period. Those 40 polls should count for 50% of the scaled weight (as opposed to 40%), and the 60 in the middle should be downgraded to only 50% (as opposed to 60%).


I haven't scoured this paper, nor the accusatorial one, but I wonder if they accounted for this.

 

[Bjorn]

I haven't scoured this paper, nor the accusatorial one, but I wonder if they accounted for this.

You should - it's an interesting read.

The senate and presidential races were both questions on a single exit poll survey. If Bush supporters were refusing to fill out this survey as hypothesized, the accuracy of the exit poll should have been just as poor in the senate races as it was in the presidential race. The presidential and senate poll results derive from exactly the same responders.

In 32 states, senate elections took place on the same ballot with the presidential race. The exit polls were more accurate for senate races than for the presidential race, including states where a Republican senator eventually won (pages 19-24).
The senate polls were significantly more accurate: paired t-test, t(30) = -2.48, p < .02, if outlier North Dakota is excluded. Therefore the Mitofsky/Edison hypothesis of reluctant Bush poll responders is irrelevant to explain the discrepancies between the exit poll and election results in the presidential race.

This seems to cast some doubt about your theory.