From Nature magazine via slashdot: a survey by experts of articles on 42 science related topics (e.g. "Cambrian Explosion," "lymphocyte," "neural network," "quark," etc. Complete list here) found that the Wikipedia articles contained an average 4 errors per article, whereas Encyclopedia Brittanica contained on average 3. Each encyclopedia contained four instances of "serious errors, such as misinterpretations of important concepts" in the 42 articles surveyed. The rest were minor factual errors. Interestingly, the article with the most errors seems to have been the same for both Wikipedia and Brittanica: that on Dmitry Mendeleev, where Wikipedia contained 19 errors, and Brittanica 8, according to Princeton Professor Michael Gordin who recently wrote a book on Mendeleev.
Two further facts are of note: the reviewers said that the Wikipedia articles were usually more confusing and harder to follow, and a Wikipedia enthusiast has noted that the comparison may be unfair since wikipedia articles are much longer on average (2.6 times longer, to be precise), and so actually have fewer errors per kilobyte of data.
A similar study, though of much smaller scale, was undertaken by Ed Felten some time ago, which I discussed at the time here.
Posted by Kenny at December 15, 2005 3:15 PM| Trackbacks |
TrackBack URL for this entry: https://blog.kennypearce.net/admin/mt-tb.cgi/148
|