Wikipedia, the encyclopaedia that relies on volunteers to pen nearly 4 million articles, is about as accurate in covering scientific topics as Encyclopaedia Britannica, the journal Nature wrote in an online article.
The finding, based on a side-by-side comparison of articles covering a broad swath of the scientific spectrum, comes as Wikipedia faces criticism over the accuracy of some of its entries.
Two weeks ago prominent journalist John Seigenthaler, the former publisher of the Tennessean newspaper and founding editorial director of USA Today, revealed that a Wikipedia entry that ran for four months had incorrectly named him as a long-time suspect in the assassinations of president John F. Kennedy and his brother Robert.
Such errors appear to be the exception rather than the rule, Nature said in Wednesday's online article, which the scientific journal said was the first to use peer review to compare Wikipedia to Britannica. Based on 42 articles reviewed by experts, the average scientific entry in Wikipedia contained four errors or omissions, while Britannica had three.
Jimmy Wales, who founded St. Petersburg, Florida-based Wikipedia in 2001, said the accuracy of his project varies by topic, with strong suits including pop culture and contemporary technology. That's because Wikipedia's stable of dedicated volunteers tend to have more collective expertise in such areas, he said.
Next month, Wikipedia plans to begin testing a new mechanism for reviewing the accuracy of its articles. The group also is working on ways to make its review process easier to use by people who have less familiarity with computers and the Internet.
Unlike Britannica, which charges for its content and pays a staff of experts to research and write its articles, Wikipedia gives away its content for free and allows anyone - amateur or professional, expert or novice - to submit and edit entries.
Wikipedia, which boasts 3.7 million articles in 200 languages, is the 37th most visited Web site on the Internet, according to the research service Alexa.
