Reply #38: Use comparison of "critical" vs. "non-critical" states. [View All]

[mountebank]

That is found here:

http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=36314&mesg_id=36314

In a nutshell: Critical states showed twice as much movement from exit polls to vote tallies than non-critical states. This despite the critical states having an average sample size of almost double that of non-critical states, which would usually mean more accurate results.

We definitely need to put a lid on the analysis of statewide exit polls between so-called "paper" and "non-paper ballot" states. This analysis is getting repeated in a lot of places, including media outlets (truthout, etc.) and is seriously flawed. For one, states with a paper ballot were included in the non-paper ballot category. Second, a second analysis showed no association, per:

http://www.bluelemur.com/index.php?p=405

But the difference between swing states and non-swing states is suggestive.