One of the GREEAT things about descriptive statistics is ZERO margin of error. This avoids all the debates about polling accuracy, methods, etc. PLUS, zero margin of error is just that!!
With OHIO punch card ballots (3/4 of the vote) and no precinct marks on the ballots, switching ballots from a precinct where the punch was a Kerry vote to a precinct where that punch counts as a Bush vote is only detectable using descriptive statistical analysis of the results. Recounting does nothing. And, guess what, there is a high percentage shift evidenced in this analysis, such that precinct ballot swapping seems the only explanation:
The 2004 Ohio Presidential Election: Cuyahoga County Analysis
How Kerry Votes Were Switched to Bush Votes
http://jqjacobs.net/politics/ohio.htmlQUOTES:
"In a subset of 166,953 votes, one of every 34 Ohio voters, the Kerry-Bush margin
shifts 6.15% when the population is sorted by outcomes of wrong-precinct voting."
"Wrong-precinct voting can occur in several ways. Voters can use their own precinct's ballot at another precinct's voting machine. Additionally, the outcome can be altered if ballots are switched to a different precinct after the voting and before the counting. This study deals with the outcome of and evidence of cross-voting and does not determine which of these methods altered the results. Wrong-precinct voting and ballot switching have the same effect, and herein cross-voting and vote-switching refers to both possibilities."
"I define "vote-switching" as major candidate cross-voting. One major candidate cross-vote changes the election margin by two votes; as one major candidate loses a vote the other gains the cross-vote. Vote-switching is distinguished from cross-voting because impact on results varies depending on for which candidate cross-votes are counted. Vote-switching results when the two major candidates are collocated in the same ballot order position, either from ballot switching between such precincts or voters cross-voting at such locations."
MUCH MORE .............