When Nature Magazine published a study which suggested Wikipedia is far more reliable than is commonly believed, fist-waving librarians were all over the Slashdot scene with their criticisms. The study gave reviewers a blind test to examine a parallel sample of articles from Wikipedia and Britannica, and demonstrated that the average number of errors in a typical Wikipedia science article, which was 3.86, is not substantially more than in Britannica, which had 2.92. Not every error is equally erroneous in the study, but that reflects poorly on Britannica regardless. It meant the most renowned encyclopedia for entry-level reference work is seldom more accurate than the anyone-can-edit-“populist history of the world,” as editors at Britannica call it.
And even though Wikipedia contained an average 0.94 more inaccuracies per article, consider further that it contains 1.4 million articles in the English version alone, while Britannica contains a mere 120,000. If each wiki entry contains the same average error amount, it must be admitted that Wikipedia simply holds more factual information than Britannica. In fact, for Wikipedia to contain fewer accuracies than Britannica it would need more than 25 average errors per article. It’s unempirical to say that Wikipedia is an unreliable source of information. It might even be unscientific to say that, since after all, Nature conducted the study.
Wikipedia founder Jimmy Wales has always maintained that The Wiki and its community are built around a self-policing and self-cleaning nature which helps to ensure its accuracy. Of course, vandalism does happen on Wikipedia. But the users have dozens of metawiki tools to check users and watch for errors. The more advanced editors use bots, for example, which allow users to sort thousands of articles a minute, making corrections as the bot moves from article to article. Not just anyone can get their hands on a bot, however, because the peer-review process is the law in world of wikis.
To get your bot approved, one has to be a reputable contributor first. After you run it through some tests, other users discuss the bot on the talk page and decide whether you can use it on Wikipedia or not. The peer-review community is why, when comedian Stephen Colbert told his thousands of viewers to change the article on Elephants to say “the number of elephants has tripled in the last six months,” the errors were quickly found and the article was locked to prevent further vandalism.
In politically sensitive areas such as climate change where the NPOV (neutral point of view) is disputed, contributors have had to battle with skeptics pushing a POV that is out of kilter with mainstream scientific thinking. But this usually requires no more than a little patience. Wikipedia’s users are generally interested in the reasoning behind proposed changes to articles. Backing up a claim with a peer-reviewed reference makes a world of difference. And every edit ever made to an article is archived, which allows everyone to see what changes have been made, by whom they were made, and the reasons for doing so.
The MediaWiki Foundation developed wiki technology with the intention of keeping it free and open-source. “Because ideas want to be free,” the MediaWiki logo says. But ever since Wikipedia was launched in 2001, the proprietary encyclopedias began losing their dominance to the open-source community. Microsoft, in an attempt to recapture the encyclopedia market, announced it will be using wiki-like features in its next release of Encarta. All proposed changes to Encarta, however, will be reviewed by a small staff of editors but still wouldn’t keep up with Wikipedia’s exponential pace. Microsoft doesn’t understand that wiki communities are not isolated individuals whom it can easily exploit and manipulate. To run a wiki requires committed people who genuinely care about the community for it to work. Why should anyone pay money to help edit Encarta articles so that Microsoft can become richer? I would much rather edit a Wikipedia article and help make the world more free and open-source.