From my conversations with many of the new students, I know that one of the biggest concerns of first year grad students is the stats course required of them this fall, Pols 800. I remember when I started that I was both worried about the class and annoyed that I had to take it – I was going to be a political scientist, I thought, I should be reading treaties and history books, not doing equations!
As you will soon see, that “scientist” part of political scientist is taken quite literally, and the discipline has adopted the data analysis tools used by our nearest cousins in economics and psychology to try to explore the world around us and identify general principles that can be used to explain various kinds of human behavior. Simply put, it is impossible to do political science without a good understanding of statistics. However, it turns out that it’s not really all that hard (at least for the basics) and once you get into working with data statistics actually turn out to be pretty interesting.
To help you get started, I want to recommend two great resources. The first is Khan Academy. I’ve been talking this website to everybody I know – it’s one of the best things on the web. This is the link to Khan’s series of videos on statistics. These videos cover everything you will learn in 800 through mid-November or so, and they are awesome. Khan’s genius is that he is empathetic with new learners. He doesn’t just present information and say ‘take it or leave it,’ he is capable of imagining what a person new to the field is going to have trouble with and walks you through each step with no skipping over the stuff he thinks is boring or should be obvious. (Those of you teaching recitations should take note! My take is this is the single most important attribute of a good teacher).
The other resource is a blog post that went up today on the blog of Marc Bellemare, a professor at Duke who works on development issues. In the post are links to pdfs he has written describing linear regression (the work horse of political science) and about causation, a surprisingly tricky subject. He also included links to some other interesting sources on causation issues.
Check them out – they will save you a ton of work and pain in the long run.