Reply #3: Already done to some degree; Hout, Freeman, myself, etc. showed

[berniew1]

statistically that there was a big unexplained swing in touchscreen counties in Florida, Ohio, Pennsylavania, etc.
Link to all these studies in
http://www.flcv.com/fraudpat.html

my study for Florida which got similar results to Houts(U. of Calif.-Berkeley) study showed big unexplained swing in the big touchscreen counties.
The documentation of widespread vote machine fraud patterns and other manipulation of votes in those counties are seen to explain this swing. See above URL.

The best Exit Poll data by county/precinct has not been released, but the big swings are in states with documented touchscreen switching like Florida, Ohio, Pennsylvania.
http://www.flcv.com/ussumall.html


my statistical vote analysis for Florida was
http://www.flcv.com/fla04EAS.html