Designing for commitment in online communities

Robert Kraut has tried to find the factors influencing participants’ commitment to online communities. You can see the links to Kraut in Nicola Avery’s comment in my former post. Thanks to Nicola. Now I try to use some concepts and assess the results Kraut presented in his articles.

Is it possible to design for commitment in online communities? – a fascinating question. Hooking the newcomers – how can it be implemented? How to manage participation, do we know the rules of thumb for good practitioners? Some results seem to be clear: the quality of relationship is critical to the willingness to stay. Newbies must not be left alone, somebody has to respond and personalize the interaction. It is better to speak about us, not to you. This interaction must happen quickly. The types of early interaction which influence on how much newcomers contribute: best was a task request, second positive feedback, then assistance, critics or comments and less influencing was social chat.  Institutionalized vs personalized welcomes: it is easy to guess the results.

Identity-based commitment and bond-based commitment – these concepts I want to ponder more. Identity -based commitment = care about ideas, content, not people (for instance wikipedia workers). Bond-based commitment means social relations between people (Facebook). These two commitments can be antagonized, compete with each other – or follow each other, move from bond to identity based. I remember when CCK09 began: I had decided not to participate the course, but when I saw the names of Finnish participants and read their comments in Moodle, I wanted to come in.

Kraut says that to increase identity-based commitment you have to emphasize subgroups for interests and member profiles must be found. Bond-based commitment happens between individuals, private relationships. He also used concepts identity-inspired and bond-inspired. This is near our first researches about CCK08 (Jenny, Roy and John) – they asked if we were focused in concepts and theories or people. I still remember that question. Kraut says that identity-based orientation is stronger and interpersonal new bonds take longer time to develop.

Agent Based Models to conduct virtual design experiments: not an easy job to do, is it possible at all? The logic of social life is at odds with the logic of design. Theories of social psychology are too narrow to understand multidimensional phenomena.  Anyway, experiments can give some information about them. Moderating discussion: none, community level or personalized, which is best? Personalized moderation works best, but when the number of topics increases, results are not clear any more.

What did I learn? Newcomers need feedback and it should be personalized. When a participant has self-selected commitment from the very beginning, he probably continues and finds the way. I knew this all, I have a long history as online teacher – and I was pleased to see my beliefs supported by research results. It was a good mirror.

All depends.. In PLENK we have broad contents varying every week. Participants’ interests are different.. I have to write another post about it. Kraut did not speak about networks or crowds but human behavior has psychological laws  – therefore scientific basic knowledge is valuable. – After writing this I read Ian Wood’s blog post about PLENK (Chris Jobling tweeted it) and .. Ian seems to know everything about online communities, it is learning by doing and participating 🙂

Nicola writes about “we and you” in one of her blogs. The web and people are not that simple, I agree. Nicola concludes: a person who tries to tell me what to do without knowing me is capable of causing harm even with good intentions – so I am better off avoiding the exposure. Heli continues: Advices seldom work properly, they tell more about givers than receivers.