Hugh Pickens writes "Presidential candidate Mitt Romney received eight more votes than candidate Rick Santorum or 0.007 percent of the total number of caucus votes in the Iowa caucus, 'eking out a victory' on the path to winning the Republican nomination for president but experts in statistics say Romney and Santorum actually tied. 'From a statistical point of view, you can't say Romney won any more than you can say Santorum won,' says Charles Seife, a professor of journalism at New York University who studies election error. That's because in the Iowa caucus, where voters marked their choices with check marks or by writing the candidates' names in by hand, the error rate in counting the votes, which is also done by hand is orders of magnitude above the victory margin — around 0.5 to 1 percent. There are several sources of error that could easily render eight votes meaningless." (Read on for more.)Hugh Pickens continues: "First, ballots sometimes stick to the bottom of ballot boxes when the boxes are overturned, and fail to be counted. Next, election officials occasionally misread messy handwriting, or tally their totals incorrectly. Finally officials can misjudge who a voter intended to vote for: 'You'd be surprised how often people place a check mark in an ambiguous place,' says Seife. Whether it's statistically significant or not, any official declaration of victory can have big ramifications. With political pundits regarding Romney's 'victory' as evidence that he's in a good position to win the Republican nomination, the failure to recognize a statistical tie in Iowa could impact the future of the country. 'It's Romney, not Santorum, who can head to New Hampshire claiming the win,' writes Nick Rizzo. 'But if you just counted the exact same votes all over again, there's a good chance the result would be different.'"