Thanks all for the input on content analysis. I also find Krippendorff's book and ideas quite helpful. Most thankfully, Andrew Hayes at Ohio State has created a handy SPSS macro for calculating Krippendorff's Alpha, complete with a bootstrapping algorithm for estimating the confidence intervals. See: http://www.comm.ohio-state.edu/ahayes/SPSS%20programs/kalpha.htm If you have SPSS, my understanding is that this is the only Krippendorff's Alpha solution you'll need. Be warned, though: run it once without the bootstrapping calculation just to get the calculation first, because WHOAH does it take a lot of processing power. Basically, if you're doing the full 10,000 bootstrap, run it right before a trip to the coffee shop, and even then, it may still be running when you come back. (You can still email or whatever while it runs in the background.) On a somewhat less intimidating note, I don't think anybody's mentioned the book ANALYZING MEDIA MESSAGES, by Riffe, Lacy, and Fico, which is pretty accessible and still a good on-ramp toward more advanced work. I like KK's book better; if you want to understand the theory behind calculating reliability estimates, for instance, there's no comparison. Pick Krippendorff. Yet I'd be much more comfortable assigning the Riffe book for an undergrad methods course, for instance. Cheers, Bill Bill D. Herman Ph.D. Candidate Annenberg School for Communication University of Pennsylvania bherman@asc.upenn.edu