WIKIPEDIA - MORE accurate than Britannica

[iconoclastNYC]

Slashdot:

Nature magazine recently conducted a head-to-head competition between Wikipedia and Britannica, having experts compare 42 science-related articles. The result was that Wikipedia had about 4 errors per article, while Britannica had about 3. However, a pair of endevouring Wikipedians dug a little deeper and discovered that the Wikipedia articles in the sample were, on average, 2.6 times longer than Britannica's - meaning Wikipedia has an error rate far less than Britannica's."

http://en.wikipedia.org/wiki/Wikipedia:Village_pump_(news)#Nature_follow-up:__How_do_the_article_sizes_compare.3F

http://news.bbc.co.uk/1/hi/technology/4530930.stm

[Captain Hilts]

1. Pick a controversial topic, or a biography, and accuracy is lousy.

On some entries I can recognize the zealots who are editing the entry.

It's just not reliable.

[iconoclastNYC]

3. Well this study says otherwise.

But whatever, your oppinion, with nothing to back it up is more imporant.

[Iris]

5. Does this meet your approval?

http://www.nytimes.com/2005/12/04/weekinreview/04seelye.html?ex=1134795600&en=fb451049294cb841&ei=5070


Snared in the Web of a Wikipedia Liar

FALSE WITNESS How true are "facts" online?


By KATHARINE Q. SEELYE
Published: December 4, 2005

ACCORDING to Wikipedia, the online encyclopedia, John Seigenthaler Sr. is 78 years old and the former editor of The Tennessean in Nashville. But is that information, or anything else in Mr. Seigenthaler's biography, true?

The question arises because Mr. Seigenthaler recently read about himself on Wikipedia and was shocked to learn that he "was thought to have been directly involved in the Kennedy assassinations of both John and his brother Bobby."

"Nothing was ever proven," the biography added.

Mr. Seigenthaler discovered that the false information had been on the site for several months and that an unknown number of people had read it, and possibly posted it on or linked it to other sites.

If any assassination was going on, Mr. Seigenthaler (who is 78 and did edit The Tennessean) wrote last week in an op-ed article in USA Today, it was of his character.

The case triggered extensive debate on the Internet over the value and reliability of Wikipedia, and more broadly, over the nature of online information.

Wikipedia is a kind of collective brain, a repository of knowledge, maintained on servers in various countries and built by anyone in the world with a computer and an Internet connection who wants to share knowledge about a subject. Literally hundreds of thousands of people have written Wikipedia entries.

[Captain Hilts]

7. On some subjects that might be true.

But on others, people with agendas are changing the entries every day to serve their views. Imagine the folks that call in to C-Span's Washington Journal making a political encyclopedia and there you go.

It is inaccurate in many cases.

And spare me the snarky attitude.

[Captain Hilts]

8. The study pertains only to topics relating to science.

Where it is possible to deal objectively with facts.

Get off the science topics and venture into history, and anything goes.

[mainegreen]

2. WIKI is a fantistic source for any technical info.

Its unreal the amount of great cross-linked data on hardware/software/tech info there.

[alcibiades_mystery]

4. Endevouring? LOL

Endeavoring, I should think.

Never mind. With a thousand eyes all misspellings are shallow...

[Liberal In Texas]

9. British spelling. Nature is a British magazine.

And they contend that they way they spell things is the CORRECT way to spell things in ENGLISH.

LOL to you.

[Lisa]

6. actually, it's important to state that it's "error rate per # words" ...

Or something like that. Just saying "error rate" is rather vague. From a researcher's standpoint, if an error is a serious one, it will be a major mistake whether it's in an article that's 5 sentences long, or a book that's 500 pages long. Typos are one thing, but major statistical or philosophical omissions (or misattributions, or erroneous statements, which can be even worse) can cause problems regardless of the article length. (Not to point any fingers here ... I've seen these in refereed journals and edited books, not just Wiki entries.)

So some provision ought to be made for how serious the error is ... and unfortunately, there's no shortcut for measuring that (different for each situation).