Thursday, December 15, 2005

Wikipedia is as accurate as Britannica

Since we have referenced Wikipedia a few times, and it is a great example of Open Source development (being licensed under the GPL) I thought this was interesting. Advocates of Open Source argue that many eyes and many hands catches more bugs. While detractors say that with no barrier to entry you end up with unqualified contributors. This is especially tough for Wikipedia since anyone with a web browser can change articles and have their changes published immediately. Well, Nature decided to settle the argument by doing a peer review comparison between Wikipedia and Britannica articles on various scientific subjects.

CNN is reporting on Nature's report

The results: the free and amateur built Wikipedia is as accurate as the commercial and professionally built Britannica.

Based on 42 articles on science reviewed by experts where the experts didn't know if the article came from Wikipedia or Britannica. The result was that Wikipedia had on average 4 errors or omissions while Britannica had 3. In some articles Wikipedia had less then Britannica. On average though the Wikipedia articles were longer and covered more material. If you take the length into account then they were close to the same.

Subject Tags: [] [] [] [] [] [] []

2 comments:

Craig said...

Taking length into account Wikipedia articles have less than half as many errors per byte than EB articles, based on the data from the Nature study. Another way to view this is that on average Wikipedia articles have significantly more correct things than EB articles.

OTOH, encyclopedias are summaries of existing research, not research publications themselves. Hence, no encyclopedia — not EB or Wikipedia or any other — is really an authoritative source for citation. A good encyclopedia like Wikipedia encourages citing of sources within an article, however, so that interested readers can look up the original research.

Disclaimer: I write for Wikipedia, so I might just be biased.

unused said...

I find it unlikely you are biased. All Wikipedia writers hold strongly to a NPOV.