Wikipedia articles that are controlled by a small group of editors that coordinate their work closely seem to be of higher quality. This is one of the main suggestions made by Kraut an Kittur, whose work was previously featured on this blog, in a paper at the CHI conference in Boston this spring. The paper analyzes the impact number of editors and editorial concentration (ie, inequality of contribution) have on independent or interdependent tasks and on article quality. Their working hypothesis was that for interdependent tasks, especially if they are to produce high quality outcomes, it is preferred to have a smaller core of high impact editors. The results are partially supported by the data.
The success of Wikipedia has demonstrated the power of peer production in knowledge building. However, unlike many other examples of collective intelligence, tasks in Wikipedia can be deeply interdependent and may incur high coordination costs among editors. Increasing the number of editors increases the resources available to the system, but it also raises the costs of coordination. This suggests that the dependencies of tasks in Wikipedia may determine whether they benefit from increasing the number of editors involved. Specifically, we hypothesize that adding editors may benefit low-coordination tasks but have negative consequences for tasks requiring a high degree of coordination. Furthermore, concentrating the work to reduce coordination dependencies should enable more efficient work by many editors. Analyses of both article ratings and article review comments provide support for both hypotheses. These results suggest ways to better harness the efforts of many editors in social collaborative systems involving high coordination tasks.
The research is important because it pricks to a certain extent the Wisdom of Crowds ideology, which for a time has lured many to believe that Wikipedia is the product of unaided, uncoordinated, crowdsourcing. (Although I usually do not link to Wikipedia articles as reference sources, this time I do it to reveal how the Wikipedia community thinks about concepts such as “crowdsourcing”.) For more infromation about Wikipedia, see Andrew Lih’s recent book, the Wikipedia Revolution.
Related articles by Zemanta
- How to Architect For Web 2.0 – Three Postulates and Ten Must-Dos (web2.sys-con.com)
- LinkLog: Layered Intelligence (dorai.wordpress.com)
- Wikipedia springs free labor leak (theregister.co.uk)
- How to Architect For Web 2.0 – Three Postulates and Ten Must-Dos (apache.sys-con.com)