I'm working on this big research project spanning four U.S. presidential elections and I'm hip deep in that tedious yet nerdishly appealing stage of writing scripts to extract data, recode it, and run initial analyses. If you have no idea what that means, you're among the few, the lucky few.
I have a basic analysis of only 2000. I'm working on the same analysis of 2004. It's damn slow going -- and that's by someone who is really good at SPSS and wrestling with large datasets. Do not try this at home. No, really. Don't. UGA has a neat virtual lab setup that includes accessing SPSS from your home box, but after test driving the thing with a modestly-sized set of data, it's my recommendation you do not try this at home. Too slow, too clunky. Reminds me of my days on a 300-baud modem accessing the mainframe.