Thanks for the great article.

One theme that we are working on in a different context - the NSDL Developmental Math Collection and a Knowledge Exchange Network for Dev Math instructors - is a social and technological infrastructure to explicitly offer users the opportunity for deeper levels of engagement.We are looking for ways to encourage faculty users to move from commenting on an article in a compendium of exemplary practices to working with the loosely-coupled team that maintains the article to working on a more tightly-coupled team to test adaptations of a practice in a new context, etc.

We think that we need people with roles that go beyond the typical 'online mayor' role of welcoming newbies to explicitly facilitate deeper and ongoing involvements - this is obviously a cone-shaped or funnel type of demographic, where only a subset of the community will consistently engage in more longer term collaborations. At the community scale we envision, it does not seem that we need advanced technology to help identify candidates for further nurturing as community members, but maybe we should be thinking more about scale-up than we are: I liked your emphasis on finding a community size that would feel right to the members. Our approach is to focus on regional scopes aggregated into a state-wide network).

 So my question is whether you have explicit ways to do this in VIPER, i.e., Come for the Content, Stay for the (simple parts of the) Community, Gradually Move into deeper levels of engagement, etc.

Tom, You identify one of the great challenges (the great challenge?) in online communities. "I have this great group of casual participants--how can i grow a small group of dedicated participants from that?". Or, in other words, we've got a bunch of people sharing (telling stories in the barn) and critiquing (sitting around kvetching at each other in the barn); how do we get them to collaborate to build something (holding a barn raising)?

Every online community wants this, and yet, just as plainly, we're not all experiencing the success of Wikipedia or Linux. Here are a few random thoughts, which are my own, and are only partly considered sensible by the other "Pit VIPErs" in IONiC, the "leadership" council that hosts this humble website. The Long Tail is as close to a Law of Nature as we get in online systems. You mention a "funnel" of distribution; most, such as Clay Shirky (http://www.shirky.com/writings/powerlaw_weblog.html) describe the effect as a Power Law (well known to you in Mathematics, no doubt!), Zipf curve, 80/20 rule, Pareto Principle, or Chris Anderson's Long Tail. A few participants (those to the left on the x-axis) will participate a lot (a high value on the y-axis), and a lot of participants will contribute very little, the pattern resembling (with canny correlation) a curve such that y = a*EXP(-bx). Even in the most successful communities, the vast majority will contribute very little, and a small number will contribute a huge amount. Trying to change this dynamic without forcing the participants to do more (unlikely, unless you have a private army) or bribing them to do more (unlikely, unless you are their employer) won't work. There are two different strategies to deal with this dynamic.

1. Size. This is, frankly, part of how Wikipedia and Linux do it. Wikipedia is perhaps the more extreme example. It probably has more users than the drugs alcohol, tobacco, and caffeine, combined. The *overwhelming* majority never edit a comma. But by sheer scale, if you grow your community enough, you'll garner more and more people who want to create together. In a mathematical sense, you're growing "a", the pre-exponential that determines the size of the curve, not its shape.

 2. Strengthening the Community of Practice (h/t Wenger). If a community of practice is "a group that comes together in order to accomplish something", you have two sources of "glue" that anchor the CoP: the community and the practice. To build this, you either need to build the "community" bit (which is a *social* process) or the practice bit (which is an *infrastructure* matter). In other words, you have to make them want to work *together*, because they enjoy doing so, or work on *this thing*, because they really want it to happen. Of course, you can try to do both, but the first is best brought about by bringing people together (face to face!) as much as possible, letting them build strong ties, and keeping them in touch with each other. To do the second, the project has to be an exciting goal that seems plausible, and worth far more to them than the costs of participation.

Of course, these have always both been true of any volunteer organization. The Internet, however, allows for much greater Size, much better Community (it's easier to maintain those ties after your annual meeting is over) and better, more plausible practice (both by collecting the small number of people who *really* want this done, and in giving them the tools to do it remotely). These factors influence "b" in the equation-how quickly the dedicated users drift off into the pool of lurkers.

Our strategy is really to do all these at once. We have plans (far from actualized, but plans nonetheless) to continue growing the *size* of the group through continued outreach at meetings various and sundry, keeping people aware, just trying to get people involved at *any* level, while strengthening the Community of Practice by both community building through summer workshops for small groups of educators and by trying to educate people in our meeting visits and publication just how much they can get out of an online community, and therefore that it's really worth building.

Scott, you may want to check out this data reported from commercial sites about the levels of engagement. This looks like the kind of analysis we can adapt to the higher ed context.

http://blogs.forrester.com/groundswell/2010/01/conversationalists-get-o…