(Jack Vinson’s written about it as well)
The findings, while not revolutionary or completely unexpected, are interesting.
It turns out that the role of the individual leader was not a major influence:
Contrary to our own expectations, the efforts of community leaders were not directly associated with community posting or viewing activity. We conclude that community leadership affects a more foundational building block of the virtual community, while virtual community activity (such as posting or viewing) is directly affected by individual membersÃ¢â‚¬â„¢ needs (usefulness) and experience (offline interaction).
In other words, the level of activity in a virtual community is driven by the members, not the leaders: their interaction with each other and their perception of the community’s usefulness.
The community leaders are significant in establishing the fundamentals and can have influence, but need to do so indirectly, by increasing the “percieved usefulness” of the community or by creating/simulating “offline interaction.”
What the study data shows is that:
posting activity is influenced by offline interaction (p < 0.01) and that viewing activity is affected by perceived usefulness (p < 0.01). Both models were significant at p < 0.01 level (F-value =16.601 and 22.563), with the predictors explaining 31.6% and 41.0% of the total variance, respectively.
The correlation between offline interaction and posting activity “suggests that offline meetings strengthen solidarity and intimacy among community members.” This means can community leaders, and those trying to develop meaningful communities, should “expl0re integrated ways to strengthen the social identity of their community members” across the online/offline divide.
For example, when face-to-face meetings aren’t possible, using photo avatars, audio/video responses, and real names / profiles can make the community feel more like an offline community. Of course this won’t work for all communities, but it can make the community feel more “real” and therefore more likely to result in contributions, not just views.
The second conclusion, on the face of it, risks turning into a tautology: people spend more time reading content of communities they percieve as useful. (Remember Beavis & Butt-Head? “I like things that don’t suck”).
Making sure that the user perception of usefulness stays high, however, is more difficult to achieve than it is to describe. The authors point to using peer evaluation (ratings), filters for obsolete or redundant material, and reward systems for valued postings as ways of increasing the percieved value of a community.
There are also some interesting notes on the impact of “Infrastructure Quality” – for example, the influence of “offline interaction” on “posting activity” is made stronger when Infrastructure Quality is low – presumably because the barrier to posting is higher, so I have to be more involved to be willing to spend the time to get over that barrier.