Florida spreadsheet and scatterplots

This has probably been posted before, but just in case:

http://ustogether.org/election04/FloridaDataStats.htm

<>

<>

.......where the Diebold machines are the 'tool', the voters are the 'operators' and the various ballot templates or races are the 'devices under test'. :evilgrin:

It would certainly help to nail down the 'uncertainty' factor when all of the numbers are analyzed. :)

Clue me in, does this say there is a huge discrepency favoring Bush in the Op-Scan?

And we thought it was going to be touch screen.

.......that the Central Tabulator was the easiest and most likely place to expect fraud. :)

The data seems to suggest that they were right once again. :toast:

that the truth is out there on the ballots? I voted opti-scan in Cali and there is a ballot.

I voted in Santa Cruz County on Mark-A-Vote Optical Scan ballots. :)

My math might be wrong here, but ustogether.org (URL above) states: "the Op-Scan voting machines register about 6% larger Republican - Democrat difference than the E-Touchscreen voting machines. This is statistically significant (p < 0.01)."

The total votes with opt-scan
3,497,834
multiplied by .06 (6%) =
209,870

The spreadsheet says republican and democratic votes for all counties are:
3,953,309 Republican
3,572,099 Democrat


Subtracting 209,870 from the republicans (to correct 6% overstatement, the assumption being they are votes that should have gone to democrats) and adding 209,870 to democrats, would mean the democrats won the vote totals.

3,743,438 Republican
3,781,969 Democrat

There you have it: Adjusting for the 6% bias, democrats would have won. Time for a recount.