Off to CHIThis year I'm showing a couple of work-in-progress posters at CHI, as well as participating in a workshop on HCI for emergencies. Irina Shklovski and I will be presenting our results of a two-year ethnography of New Orleans musicians after Katrina, looking at ways they adopted and adapted technology over the long term. The posters are more in my typical area, looking at behavior in online communities.
Taking up the mop (Poster PDF, Article PDF) reveals behaviors of Wikipedia editors that make them likely to be promoted to administrator status. There had been plenty of discussion during Requests for Adminship saying that sheer number of edits shouldn't matter, but this study makes it clearer. Edit count doesn't matter, and even writing on article talk pages doesn't help much. Instead, diversity of experience and policymaking is far more valuable to those choosing admins. The model can be used as an "admin finder" to identify editors likely to pass consensus, as a self-evaluation tool for editors considering nomination, or as a voter dashboard during the Requests for Adminship processes.
Mind your P's and Q's (Poster PDF, Article PDF) looks at the effectiveness of politeness (or rudeness) at getting a reply in many kinds of online discussion groups. We found that polite language tripled reply counts in some health support groups, but rudeness was far more effective in others.
If you'll be at CHI, email or text me.