I don't have a specific reference for you, just some comments. :) I think perhaps your original claim is too strong--on both counts. Including the word "can" as you did in your reply is important. Wikis CAN lead to high-quality content. And CAN lead to active communities. (Clearly there are examples of active communities and examples of wikis that never get used much.) I am not convinced that the technology itself leads to high-quality articles in Wikipedia, and I don't think that the Nature article supports that claim in any way. One could hypothesize about characteristics of wiki that help support the goal of producing high-quality articles and try to empirically verify the claim, but to my knowledge no one has done so in any systematic fashion.
On 8/29/07, Desilets, Alain Alain.Desilets@nrc-cnrc.gc.ca wrote:
Alain Désilets wrote:
I need a good solid reference to substantiate the following claim: "Besides leading to high quality content, wikis have been shown to be good toolsfor fostering the emergence of active communities"
Does anyone know of a good research paper that looks specifically at this kind ofimpact of wikis?
Brian replied:
I think the first part of your claim needs to be substantiated first! Almost all of thecontent on the English Wikipedia, for example, is of low quality.
http://en.wikipedia.org/wiki/Wikipedia:Version_1.0_Editorial_Team/IndexI already have a solid reference to substantiate the fact that Wikipedia has quality content, i.e. the Nature Magazne study of 2005:
http://www.nature.com/news/2005/051212/full/438900a.html
This study found that the quality of Wikipedia content was comparable to that of Britannica (on a sample of 42 pages in various scientific domains).
The page you quote above does not constitute a scientific assessment of the quality of WikiPedia pages. For one thing, it's not a random sample. These are pages that people on the editorial board selected for review, and it could be that they naturally focused on pages that were a-priori more likely to present quality issues. Second, it does not compare the quality of wikipedia pages to the quality of corresponding pages on more traditional resources.
The Nature study on the other hand was scientifically sound on both of those points. It did a random sample of pages on a broad range of scientific topics. And it compared the quality of those pages to comparable pages on Brittanica.
The Nature study is also consistent with my personal experience (and that of most people I talk to) as a user of Wikipedia. In other words, whenever I go to a Wikipedia page on a topic that I know well, I find it to be good quality and accurate.
So... Until someone shows me a scientifically sound demonstration that Wikipedia pages are of significantly lower quality than other more traditional resources, I will continue claiming that wikis like Wikipedia can lead to quality content.
Alain Désilets, National Research Council of Canada Chair, WikiSym 2007
2007 International Symposium on Wikis Wikis at Work in the World: Open, Organic, Participatory Media for the 21st Century
http://www.wikisym.org/ws2007/
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org http://lists.wikimedia.org/mailman/listinfo/wiki-research-l