Nature has long been linking to Wikipedia ? ?the free encyclopedia that anyone can edit? ? from its online news stories. On Aug. 25, 2005, Nature cited the Web site in a print editorial. So there was perhaps a collective sigh of relief in the journal?s offices when an in-house investigation, published on December 15, found that the site?s scientific content is not much less trustworthy than that found in tried-and-tested print encyclopedias.
Since its launch in 2001, Wikipedia has grown to 900,000 articles on the English version of the site alone. According to Alexa Web traffic rankings, it is now the world?s 32nd busiest Web site. Nature is not alone in citing it; major news outlets including The New York Times, The Economist, the BBC, and The Washington Post regularly do so, as does The Scientist.
?It seemed like a good time to subject it to [some sort of] peer review,? says Mark Peplow, a Nature journalist who worked on the investigation. Nature sent out pairs of entries ? one from Wikipedia and one from its print rival Encyclopaedia Britannica ? on a wide range of scientific subjects to experts for comparison. Wikipedia averaged four inaccuracies per entry, compared to Encyclopaedia Britannica?s three.
So, depending on which way you look at it, either Wikipedia is almost as accurate as the ?gold standard? for reference encyclopedias, or Encyclopaedia Britannica is only slightly more reliable than a free encyclopedia whose definition of itself acknowledges that ?vandalism and inaccuracy are constant problems.? Either way, it seems that there is little difference between the two sources.
Still, even Jimmy Wales, Wikipedia?s cofounder, recommends that news outlets be cautious when linking to the site. ?They should link to a specific revision [rather than an] article that?s under live development, which may be changed or vandalized at any moment,? he says.
Peplow?s colleague Jim Giles maintains that Wikipedia is ?a great place to find introductory material,? but he agrees that care is needed. ?We?re not implying that you go to Wikipedia to find out the exact weight of an electron,? he says.
So did Nature conduct the investigation in part to justify its citing of Wikipedia? ?No,? says Peplow. ?We thought our readers would find it interesting.? Certainly, Nature?s investigation also reveals that scientists, like the population at large, are increasingly drawn to Wikipedia. In addition to the peer review exercise, Nature surveyed more than 1,000 authors of research papers published in the journal. 70% of respondents had heard of Wikipedia, and 17% of those said they consulted it weekly, although only 10% contributed content. In an editorial accompanying its investigation, Nature urges more scientists to contribute to Wikipedia to further bolster its accuracy.
Wales has his own plans for improving Wikipedia?s reliability. By creating stable versions of entries deemed to have reached an acceptable level of accuracy, for example, and introducing a ten-minute delay before edits of controversial entries go live, Wales hopes to achieve his goal of matching or bettering Britannica. ?We?re not quite up to the level we?d like to be yet,? he says. ?Stuart Blackman