Nature has a special report at http://www.nature.com/news/2005/051212/full/438900a.html , detailing the results of an accuracy comparison between WP and EB. While the Wikipedia articles often contained more inaccuracies than Britannica's, they don't look at the article sizes in each case. With Maveric149's help, I did:
http://en.wikipedia.org/wiki/Wikipedia:Village_pump_%28news%29#Nature_follow...
Result: Average article size for Wikipedia: 6.80 KB; Britannica: 2.60 KB. Number of errors per 2KB for Wikipedia: 1; Britannica: 6.
Put another way: Wikipedia has 4 errors to their 3; our articles were also 2 1/2 times longer on average.
Can someone please check my math, I did this pretty fast, and was half asleep :) It's not 100% accurate, but I was only going for a ballpark estimate. Note: we copied the displayed WP text, not the edit box text, and removed the TOC, See also, references, external links, and any other big tables or lists. The WP text came from just before the Nature article was published.
Raul654 and I separately submitted stories to Slashdot, and I would suggest anyone willing do something similar. The more requests they have for this, the more likely they are to accept it.
brian0918
That's just great information.
What would be interesting also would be to draw a social map of articles in WP and EB, in order to see if they take care of the same topics or concentrate on different things.
Brian wrote:
Nature has a special report at http://www.nature.com/news/2005/051212/full/438900a.html , detailing the results of an accuracy comparison between WP and EB. While the Wikipedia articles often contained more inaccuracies than Britannica's, they don't look at the article sizes in each case. With Maveric149's help, I did:
http://en.wikipedia.org/wiki/Wikipedia:Village_pump_%28news%29#Nature_follow...
Result: Average article size for Wikipedia: 6.80 KB; Britannica: 2.60 KB. Number of errors per 2KB for Wikipedia: 1; Britannica: 6.
Put another way: Wikipedia has 4 errors to their 3; our articles were also 2 1/2 times longer on average.
Can someone please check my math, I did this pretty fast, and was half asleep :) It's not 100% accurate, but I was only going for a ballpark estimate. Note: we copied the displayed WP text, not the edit box text, and removed the TOC, See also, references, external links, and any other big tables or lists. The WP text came from just before the Nature article was published.
Raul654 and I separately submitted stories to Slashdot, and I would suggest anyone willing do something similar. The more requests they have for this, the more likely they are to accept it.
brian0918 _______________________________________________ foundation-l mailing list foundation-l@wikimedia.org http://mail.wikipedia.org/mailman/listinfo/foundation-l
Please let us know the links to the slashdot stories. Nature stated the articles were of similar length. Are you suggesting they were wrong?
Mgm
On 12/15/05, Brian brian0918@gmail.com wrote:
Nature has a special report at http://www.nature.com/news/2005/051212/full/438900a.html , detailing the results of an accuracy comparison between WP and EB. While the Wikipedia articles often contained more inaccuracies than Britannica's, they don't look at the article sizes in each case. With Maveric149's help, I did:
http://en.wikipedia.org/wiki/Wikipedia:Village_pump_%28news%29#Nature_follow...
Result: Average article size for Wikipedia: 6.80 KB; Britannica: 2.60 KB. Number of errors per 2KB for Wikipedia: 1; Britannica: 6.
Put another way: Wikipedia has 4 errors to their 3; our articles were also 2 1/2 times longer on average.
Can someone please check my math, I did this pretty fast, and was half asleep :) It's not 100% accurate, but I was only going for a ballpark estimate. Note: we copied the displayed WP text, not the edit box text, and removed the TOC, See also, references, external links, and any other big tables or lists. The WP text came from just before the Nature article was published.
Raul654 and I separately submitted stories to Slashdot, and I would suggest anyone willing do something similar. The more requests they have for this, the more likely they are to accept it.
brian0918 _______________________________________________ WikiEN-l mailing list WikiEN-l@Wikipedia.org To unsubscribe from this mailing list, visit: http://mail.wikipedia.org/mailman/listinfo/wikien-l
MacGyverMagic/Mgm wrote:
Please let us know the links to the slashdot stories. Nature stated the articles were of similar length. Are you suggesting they were wrong?
At least one person at [[Wikipedia:External peer review]] who has access to Britannica has found it impossible to find revisions of certain articles that are similar in size to their Britannica counterparts without going way back - like to 2002.
John Lee ([[User:Johnleemk]])
On 12/15/05, Brian brian0918@gmail.com wrote:
Result: Average article size for Wikipedia: 6.80 KB; Britannica: 2.60 KB.
We badly need to tighten up our writing.
On 12/15/05, Tony Sidaway f.crdfa@gmail.com wrote:
On 12/15/05, Brian brian0918@gmail.com wrote:
Result: Average article size for Wikipedia: 6.80 KB; Britannica: 2.60 KB.
We badly need to tighten up our writing.
Remember The britanica guy haveing a go at us becuase our article on encyopedia was too short? -- geni
geni wrote:
On 12/15/05, Tony Sidaway f.crdfa@gmail.com wrote:
On 12/15/05, Brian brian0918@gmail.com wrote:
Result: Average article size for Wikipedia: 6.80 KB; Britannica: 2.60 KB.
We badly need to tighten up our writing.
Remember The britanica guy haveing a go at us becuase our article on encyopedia was too short?
The "tightening up" is less about content than writing style. It's a special skill to take a piece of text and pare it down to its essentials by removing excess verbiage. Recently, I happened to look at the Wikijunior article on Chile. It's very badly written because it primarily fails to take into account that it's written for children. Apart from the evident political tone of the article I noted the seemingly innocent sentence, "In 1970, a man named Salvador Allende was elected." I have no factual dispute with this sentence. What, however, is the purpose of the phrase, "a man named"? The sentence would have read just as well without it
Tightened up writing meqans removing this kind of fluff.
Ec
On 12/15/05, Tony Sidaway f.crdfa@gmail.com wrote:
On 12/15/05, Brian brian0918@gmail.com wrote:
Result: Average article size for Wikipedia: 6.80 KB; Britannica: 2.60 KB.
We badly need to tighten up our writing.
Blame this on the ability to edit by section and the strong merge/agglomeration camp. Merging two small articles into one bigger article inevitably leads to ungainly, uneditable articles over time.
The Cunctator wrote:
On 12/15/05, Tony Sidaway f.crdfa@gmail.com wrote:
On 12/15/05, Brian brian0918@gmail.com wrote:
Result: Average article size for Wikipedia: 6.80 KB; Britannica: 2.60 KB.
We badly need to tighten up our writing.
Blame this on the ability to edit by section and the strong merge/agglomeration camp. Merging two small articles into one bigger article inevitably leads to ungainly, uneditable articles over time.
Hence the Fix Crappy Prose challenge for those who think they can in fact write.
- d.
2005/12/15, Brian brian0918@gmail.com:
Nature has a special report at http://www.nature.com/news/2005/051212/full/438900a.html , detailing the results of an accuracy comparison between WP and EB. While the Wikipedia articles often contained more inaccuracies than Britannica's, they don't look at the article sizes in each case.
In fact, it seems they did look at the size. If you look carefully at the top of the list of articles reviewed by Nature, you will find the following sentence :
All entries were chosen to be approximately the same length in both encyclopaedias.
It could even be argued that this bias in the selection actually improved Wikipedia's performance by screening out a number of poorer articles.
GL
As an aside -- are they going to tell us what they thought the errors were at any point? I bring this up not only as a point related to improvement, but it would be nice to see what exactly the, err, nature of the errors were. They chose a number of history of science related articles as well, and it would not surprise me if some of our "errors" were matters of interpretation at time, rather than specifically wrong points.
It would be ideal, of course, if we could do a thorough fact-check on each of the so-labeled articles. It would be to our benefit to show that we could ferret out mistakes quickly despite not having a full-time or even professional staff. (I've started going over a few of the biographies in this fashion personally, since they are the easiest for me to check)
FF
On 12/16/05, G L glwiki@gmail.com wrote:
2005/12/15, Brian brian0918@gmail.com:
Nature has a special report at http://www.nature.com/news/2005/051212/full/438900a.html , detailing the results of an accuracy comparison between WP and EB. While the Wikipedia articles often contained more inaccuracies than Britannica's, they don't look at the article sizes in each case.
In fact, it seems they did look at the size. If you look carefully at the top of the list of articles reviewed by Nature, you will find the following sentence :
All entries were chosen to be approximately the same length in both encyclopaedias.
It could even be argued that this bias in the selection actually improved Wikipedia's performance by screening out a number of poorer articles.
GL _______________________________________________ WikiEN-l mailing list WikiEN-l@Wikipedia.org To unsubscribe from this mailing list, visit: http://mail.wikipedia.org/mailman/listinfo/wikien-l