Friday, December 12, 2008

Corruption vs. Republican: An interesting correlation

(Cross-posted with Daily Douches)

Hey Kids,

Okay. I know this isn't a shameless, juvenile attack on some idiot douchebag (like my usual MO), but I thought this was interesting.  Hell, if I'm right here I might have to start going plural in my Daily Douche award.

USA Today has a graphic showing the most corrupt states in the US.  From the article:
On a per-capita basis, however, Illinois ranks 18th for the number of public corruption convictions the federal government has won from 1998 through 2007, according to a USA TODAY analysis of Department of Justice statistics.

Louisiana, Alaska and North Dakota all fared worse than the Land of Lincoln in that analysis.

If you look at their map and compare it to this year's election results, you notice a correlation stand emerge.  The "top-tier" corrupt states overwhelmingly went for McCain (10/16) and tend to be more Red in general.  In fact, if my numbers are correct, the 2008 Red States had an average of 3.93 (+/- 2.34 standard deviation) convictions per 10,000 people.  In contrast, Blue states had only 2.72 (+/- 1.31) convictions.   That's 44% more convictions in Red States than Blue.

Now being trained in statistics, I involuntarily react to these comparisons by saying "correlation does not equal causation!"  So I'm willing to take this with a grain of salt.  It could be that Red States just get more convictions, but given the current state of the Department of Justice, I think not. 

Friday, September 26, 2008

Just how close can this get?

Slashdot is hosting a story on some brilliant work done by Mike Sheppard (a graduate student in statistics in Michigan). Using some pretty simple math, Mike shows just how close the last two elections were. His analysis can be found here. In short he observed:

"There have been 12 Presidential elections that were decided by less than a 1% margin; meaning if less than 1% of the voters in certain states had changed their mind to the other candidate the outcome of the entire election would have been different."

The last two elections were in this group. According to his analysis, in the 2000 election, had less than 300 voters had a change of heart, the outcome would have been different. From a probabilistic standpoint, the Bush victory in 2000 practically happened by chance.

Now, I bring this up for two reasons. First, as you'll see in future posts, I've been examining how the electoral college system biases our political structure so that the religious right has a greater influence on a national level than their demographic size. Second, since many of the social issues haven't changed, this clearly demonstrates the importance of getting out every single vote in this modern political environment.

Thursday, September 25, 2008


Welcome aboard to the blog Scientia Publica (the latin for "knowledge open to all" or "science of the people"). This site is meant to be a repository for some of my personal ruminations on the current state of the world, mainly through the perspective of my training as a scientist.

That training was focused in the area of something called Cognitive Neuroscience, in which I received a doctorate from the University of California, Berkeley. So I've had a blend of training in diverse fields such as psychology, cognitive science, philosophy, statistics, biology, physics, and physiology. Along with obviously guiding my research, this broad background sculpts my perspective on the political and social landscape around me. So this is how the world looks through eyes trained to see the subtle hidden structure natural systems