Here’s an interesting graphic from the NYTimes commenting upon the Obama/Clinton demographic divide.
Knowing a bit about how these decision trees are created, I can tell you that the creators could have taken this graphic in many ways and chose these particular variables, in this particular sequence, in order to tell a distinct story.
Telling a story with data is always an ‘arts meets science’ exercise. Readers usually place an emphasis on the rational message conveyed by the data, but often miss the ‘art’ piece – or in other words – they don’t realize they’re being sold a story. This is by design and why statistics can often be misleading.
So what’s hidden here? This graphic communicates that Obama wins black voters and highly educated white voters, while Clinton carries working class whites. This sounds like a recipe for electoral disaster for Obama, but we all know that he has won more states, more delegates and the popular vote.
Part of what’s missing is Obama’s dominance over the youth vote, the influence of early voting states such as Iowa, Obama’s organizing prowess in caucus states, and the electorate’s thirst for change after 7 years of Bush. That is to say, this analysis tries to simplify something which is far too complex to be simplified.
This reminds me of Gladwell’s article where he discusses national-security expert Gregory Treverton’s theory of Puzzles and Mysteries:
Osama bin Laden’s whereabouts are a puzzle. We can’t find him because we don’t have enough information.
Mysteries (meanwhile) require judgments and the assessment of uncertainty, and the hard part is not that we have too little information but that we have too much.
Puzzles come to satisfying conclusions. Mysteries often don’t.
This graphic treats the election like a puzzle, however in actuality, I believe it’s a bit more like a mystery.
[Thanks to Nathan at flowingdata]
