Oct 17 2008
Race and Obama
Bonnie has been pointing me to some neat interdisciplinary groupblogging at UCLA on the upcoming election. Two posts I’ve liked in particular lately have been Race in a post-Obama world and Residential segregation and the race vote. The first one looks at the chilling effect his winning may have on social justice and disparities work, and the latter has some excellent charts concerning the relationship between who you live with and how you may vote. As you all are aware, people still continue to self-select where you live, even if schools and public facilities aren’t technically segregated. One of the cool graphs follows:
Of course, correlation isn’t causation, and there may be a lot more to this association (thanks Epi) but this also may be a handy quick guide to living if you prefer to live in more integrated environments, too. If there were on a per-city or smaller unit than by state…hmmm….and now I’m intrigued by this Dissimilarity Index they used for the horizontal axis. He also discusses methodology at the end of the post with some neat stats things if you’re into that.
Anywho, just saying. Hoping to get this final debate off of youtube to watch with my fellow politics-interested folks out here.
October 18th, 2008 at 2:49 am
The “Residential segregation …” link doesn’t work. (I thought I’d like to glance at the data, since from the scatter in the graph it doesn’t actually look like there’s much correlation at all. It almost looks as though the line could slope at +45 degrees instead of -45 degrees just as convincingly.)
October 19th, 2008 at 12:12 am
Blah. Link is fixed - sorry about that. based on the sharpness of the graph, I assume they used a statistical package to figure out the correlation line, though I don’t know which method was used. Probably the least-squares(?) method if I remember correctly a common method…
October 21st, 2008 at 12:27 am
I think it must be weighted by population. When you notice that states like California, Ohio, Pennsylvania, etc., are close to the regression line, while places like Nevada and Arizona aren’t, it makes the correlation somewhat more convincing.