Both The Sherman Foundation and Kottke have done a few pieces on data visualisation and presentation.
This one at the recently particularly prolific TSF, covering ReConstitution 2008, shows some very powerful and potentially powerfully misleading technologies at play.
Putting all the wizardry and funny voices aside, they’re able to present real-time statistical analysis of the debates using metrics that by many standards are indicative of very relevant/important candidate traits.
Repetition is a staple of political messages (or any message, but I was first introduced to it by Sir Humphrey in Yes (Prime?) Minister). Their software shows very clearly what the key words for each candidate’s speech are and what they’re hammering home.
Probably more misleading (though in retrospect perhaps not) is the visualisation of the length of the words each candidate used. It’s a really clever visual tool, but I don’t know that in the general case that would be a good measure of a person over the space of 30 minutes of fairly scripted talk time.
Which raises a question: If you measure an idiot by a host of tools over 8 years, do you lose the ability to gauge the effectiveness of those tools?
Regardless, I’m intrigued by these ReConstitution guys. I’m keen to see what they come up with for the current debates.