There are a set of statistical skills that are essential in evaluating many kinds of results in HCI, but it isn't always straightforward to acquire them. Courses in the Statistics Department may be more theoretical than you need, while just reading the manual for a statistics package won't provide the fundamentals. Fortunately Jake Wobbrock from the iSchool has put together a self-study tutorial on practical statistics for HCI. Even more fortunately, Jake will give two guest lectures in CSE 510 this quarter, the first on Experiment Design and experiments generally, and the second on Inferential Statistics.
The self study materials are all available at ps4hci. We also have a pre-release of a version that includes directions using the open source R system: ps4hci (R version).
Before the lecture on Experiment Design, please read the following chapter. A discussion contribution is required for this chapter (same as for the research papers).
Martin, D.W. (2000) Chapter 12: How to interpret experimental results. In Doing Psychology Experiments (5th ed.). Belmont, CA: Wadsworth, pp. 245-270. [local pdf].
For the lab exercise for this part of the course, please work through the first 4 modules in ps4hci. You should first try these on your own, but after you've answered the questions, check and correct your answers using the answer key provided in ps4hci, and turn in the corrected result. (I expect that as a result everyone will get full credit for the assignment; this is OK -- the object is to help you learn the material.) We've also set up an area ("Discussion Area for Statistics Lab") on the course discussion board to ask questions and help each other with the exercises.
The self-study guide uses two different statistics packages: JMP and SPSS. You only need to use JMP in this course however. If you want to continue on with the rest of the modules, I do recommend that you also learn SPSS. (Jake is willing to give independent study credit in spring quarter for students who want to do this.)
If you use JMP, you will need to download the system. There is a UW-wide site license for JMP here.
Alternatively, you can download a trial version of the JMP software package for the exercises. You need to first register, and then you can download the trial version. It's around 425MB, so do this where you have a fast internet connection.
Or finally, you can try the R version of the exercises if you're willing to deal with a version that's still in beta.
The ps4hci modules will be graded pass/fail (which I admit is a little weird, but since the answers are online it seems more appropriate).