Exactly. No amount of hurricanes could cause there to be an interaction between type of voting machine and the vote percentages from the last election.
A more likely culprit is a lack of variance in type of voting machine in the high end of the distribution.
If you look at the graph, there are only a few counties at the upper range, so I wonder if the results are just a statistical artifact due to range restrictions and/or a lack of variability in voting methods in counties with more democrats.
In other words, even though statistically significant, they might be violating an assumption of regression analysis.
I think the wto is the worst thing to come around in a long time. im glad there are people protesting it and having a "carnival against capitalism". i only wish i could be there to take part in the festivities.
Slashdot Mirror
Exactly. No amount of hurricanes could cause there to be an interaction between type of voting machine and the vote percentages from the last election.
A more likely culprit is a lack of variance in type of voting machine in the high end of the distribution.
If you look at the graph, there are only a few counties at the upper range, so I wonder if the results are just a statistical artifact due to range restrictions and/or a lack of variability in voting methods in counties with more democrats.
In other words, even though statistically significant, they might be violating an assumption of regression analysis.
Could it be possible that a computational approach isn't necessarily the way to go?
See, for example, the work of Mark Tilden
http://www.exhibitresearch.com/tilden/
I think the wto is the worst thing to come around in a long time. im glad there are people protesting it and having a "carnival against capitalism". i only wish i could be there to take part in the festivities.