Online communities can help people form productive relationships. Unfortunately, this potential is not always fulfilled: many communities fail, and designers don’t have a solid understanding of why. We know community activity begets activity. The trick, however, is to inspire participation in the first place. Social theories suggest methods to spark positive community participation. We carried out a field experiment that tested two such theories. We formed discussion communities around an existing movie recommendation web site, manipulating two factors: (1) similarity—we controlled how similar group members’ movie ratings were; and (2) uniqueness—we told members how their movie ratings (with respect to a discussion topic) were unique within the group. Both factors positively influenced participation. The results offer a practical success story in applying social science theory to the design of online communities.

It was identified from this paper that users enjoyed being presented unique information and were more likely to contribute when they were shown personalised unique information. For example, movie rating participants were sent weekly messages telling them how their ratings differed from others in their group relative to a discussion topic. Interaction was also discovered between uniqueness and dissimilarity. Interaction describes synergy between factors which in this case suggests that the combination of uniqueness and dissimilarity factors working together is greater than the sum of the two individual factors. Further study however is required to further understand these results.

It was also uncovered that people are more likely to contribute to discussion groups when others are dissimilar to themselves. This is interesting as social theory suggests people would be more attracted to others similar to themselves and would therefore be more likely to participate in similar groups. However, the authors discovered that people with dissimilar views had longer exchanges with one another and would banter back and forth to defend their opinions. The authors also present guidelines for using their research results to increase user contributions in online communities which makes this paper a great read for designers and interested researchers.

• MovieLens - http://movielens.umn.edu/
• Recommender systems
• Productive social relations
• Under-contribution
• Similarity and dissimilarity of group members
• Uniqueness of a member's qualification for a group task
• Multi-User Dimensions (MUDs)
• Social loafing
• User satisfaction
• User clustering algorithm
• Sociology of friendship