The CSCW session on Communities ran Tuesday mid-morning and turned up a wide variety in the three talks from social psychology to blogs to reading forums. Extended entry contains details, and I'll do a separate entry on the blogging talk.
Using Social Psychology to Motivate Contributions to Online Communities by Xiaoqing Wang, Gerard Beenen, and others from Kraut's group at CMU (and Michigan and Minnesota) (paper).
Blogging as a Social Activity: Would You Let 900 Million People Read Your Diary? by Bonnie Nardi of UC Irvine; and Diane J Schiano, Michelle Gumbrecht of Stanford.
Flash Forums and ForumReader: Navigating a New Kind of Large-scale Online Discussion by Kushal Dave, Martin Mattenberg, Michael Muller of IBM Research (paper).
This work from CMU looks at the community of reviewers at Movielens.com to learn something about social psychology, and specifically the idea of "under contribution" in online communities (also termed "social loafing" - not quite the same as lurking?). They presented two sets of studies. The first tested the theory that people do more when the benefit of the activity is more explicit. They showed an interesting graphic that suggested reinforcing both personal and collective benefit would reinforce the desired behavior. In their tests, however, they only tested one factor at a time. They asked 9000 users via email to do more reviews based on their unique contribution, the benefit to them personally, or the benefit to the community as a whole. The only one of these factors that showed a significant difference was the call to uniqueness. The speaker suggested that specific appeals to individual and group benefit as a motivation technique misses the more nuanced way people interact in such a community.
The second study looked at the high performance challenges: do people perform at higher levels when given successively higher challenges. And is there a limit to that performance. And does it matter if they are challenged as a group or as individuals. In this case. they found that groups (ten people, randomly chosen) given a challenge far outperformed individuals given the same challenge. They also found an apparent limit at the extremes of the study - that at some point it isn't possible to challenge people further.
Finally, the design and longer-term implications of this study. While communities do not tend to attach goals to participation, they suggest that goals might be something to explore if under contribution is truly a concern in the community. More importantly to me, they highlighted the importance of collective effort in communities. While specific goals might be useful in some circumstances, it seems to me that a shared sense of purpose and energy is much more important to the overall performance of the community. They also had some things to say about "social facilitation" which is probably tied up with how a community encourages its members to participate.
Flash Forums are online community forums that come together and disburse in relatively short time periods for discussion around very specific topics. The best examples are IBM's WorldJams or Slashdot, where there are thousands of posts on a given topic. An anti-examples would be traditional Usenet, where people can comment weeks, months and even years after the original discussion thread was started. IBM built ForumReader to do better visualization of how Flash Forum conversations develop. The presentation shifted to a description of how the software helps people better understand and visualize a forum conversation.