Petition [EN] Pétition [FR] Petición [ES] Petiţie [RO] Ψήφισμα [EL] Petice[CZ] Petition [DE] Petizione [IT] Petycja [PL] Petitie [NL] Petíció [HU] Petição [PT] Namninsamling [SV] Underskriftindsamling [DA] Petícia [SK] Achainí [GA] Peticija [LT] Петиция [RU] Eskaera [EU] Petskribo [EO] Petició [CA] Athchuinge [GD]
National Election Archive Project
A draft analysis of New Hampshire Democratic Primary Raw Data shows that Clinton received a higher percentage of votes when ballots were counted electronically by Diebold, and Obama received a higher percentage of votes when ballots were counted by hand. The pattern is consistent with outcome-determinative vote miscount.
Well, gotta run!
These are the differences between the Zogby poll data and the reported votes, sorted on hand counted and Diebold counted:
(click for larger)
Clintons votes in Diebold counted has a quite larger difference then anything else.
National Election Archive Project also checked pairs of likesized precincts in the same county, and looked at differences in Diebold and hand counted votes for Obama and Clinton. Choosing pairs of two likesized precincts in the same county, you get pairs where the demograpic differences should not be that large. Within the pair that is.
Blue is Diebolds effect on Clinton - mostly positive - and red is on Obama - mostly negative.
Last, but not least:
Two-Tailed Difference of Means Tests by Ron Baiman This analysis is of a sample of randomly selected 36 comparably-sized machine and hand count precincts from the "middle of the size distribution" obtained by lopping off the smallest and largest precincts respectively until equal size samples (36 precincts each) with almost identical average size (762.2 votes cast for machine, 764.9 votes cast for hand counted precincts) remained in the sample. The results of a two-tailed difference of means tests (machine count versus hand count for same candidate) for Obama, Clinton, and Edwards are striking. In the standard test (assuming both machine and hand counts precincts are randomly selected from the precinct population): Odds for the results for Obama, are one out of 437 Odds for the results for Clinton are one out of 1,965 Odds for the results for Edwards are one out of 12 Both the Obama and Clinton differences between their machine and hand counts are highly significant at well above 99% level. Edwards' result is significant at the 90% level.
This analysis is of a sample of randomly selected 36 comparably-sized machine and hand count precincts from the "middle of the size distribution" obtained by lopping off the smallest and largest precincts respectively until equal size samples (36 precincts each) with almost identical average size (762.2 votes cast for machine, 764.9 votes cast for hand counted precincts) remained in the sample.
The results of a two-tailed difference of means tests (machine count versus hand count for same candidate) for Obama, Clinton, and Edwards are striking.
In the standard test (assuming both machine and hand counts precincts are randomly selected from the precinct population):
Odds for the results for Obama, are one out of 437 Odds for the results for Clinton are one out of 1,965 Odds for the results for Edwards are one out of 12
Both the Obama and Clinton differences between their machine and hand counts are highly significant at well above 99% level. Edwards' result is significant at the 90% level.
This all leading to their conclusion:
We can therefore state that the pattern of hand and machine counts is consistent with switching votes cast for Obama to votes counted for Clinton by the machines.
Do read the whole thing, it is just 4 pages.
However, several others are off by more than the statistical significance even in the hand-counted districts, including Clinton.
Choosing pairs of two likesized precincts in the same county, you get pairs where the demograpic differences should not be that large.
That's a definitely false assumption. Almost all counties include urban, suburban, smalltown and rural precints.
As for the linear fits: the odds of the result are less interesting than the odds of the linear part of the result, and they don't even give an R-squared. (It seems to me a superposition of a ringing-off sinusoid and a linear would be amuch better fit for Clinton.) I should again re-state the point that there is no control for systematic errors. *Traitor*, n. A benighted individual who perceives an illusory distinction between serving his nation and abetting the criminals who govern it.
Should have started with a disclaimer of statistics not being my best subject...
