[Wikipedia-l] Re: Wikipedia spanks Encarta, Brockhaus

[Andre Engels]

On Sat, 2 Oct 2004 23:31:25 +0100 Rowan Collins <rowan.collins at gmail.com> 
wrote:

>Hmm, something that strikes me looking at those results is that on
>several categories Wikipedia seems to do worse on the "easy" topics
>but better on the "hard" ones. I don't know if I'm just imagining it,
>and it could just be a coincidence, but that seems like an interesting
>finding (were there any graphs in the article? one could probably
>construct a graph that demonstrated patterns like that).

I have done a count for it, and it doesn't seem that Wikipedia scores 
consistently worse on 'easy' topics. In fact, Wikipedia scores about equally 
(average about 3.5) on all three difficulties. Brockhaus scores similarly in 
the easy and hard categories, but worse in the intermediate one. Encarta is 
equal to Brockhaus in the easy and intermediate categories, but failing 
clearly on the hard subjects - especially because of 6 zeroes.

Here are the numbers (for each encyclopedia and difficulty category the 
number of 0/1/2/3/4/5 scores):

EP easy: 1/2/0/5/9/5, total 78 (average 3.5)
EP medium: 4/0/2/4/8/4, total 68 (average 3.1)
EP hard: 6/2/1/6/4/3, total 50 (average 2.3)

BH easy: 0/0/4/6/6/6, total 84 (average 3.8)
BH medium: 4/0/4/5/5/4, total 63 (average 2.9)
BH hard: 1/0/3/7/5/6, total 77 (average 3.5)

WP easy: 1/1/5/2/3/10, total 82 (average 3.7)
WP medium: 1/0/4/4/6/7, total 79 (average 3.6)
WP hard: 2/0/1/7/6/6, total 77 (average 3.5)

Still, there is _something_ true in what you say. Wikipedia's scores on easy 
subjects seem to be more spread out than on the medium and hard subjects. 
There are more 5s but also more 2s. On the other hand, Wikipedia's scores on 
hard subjects show a general content of quality - only one 2 and no 1s among 
20 subjects that at least had something.

Maybe it has something to do with the Wiki method of working on on one 
another's texts. Maybe the more difficult texts are mostly work of one or 
two people, with the rest only acting as a copy-editor. Apparently those are 
giving good work. On the easier subjects, 'everyone' thinks he/she can write 
something, which can have both positive and negative effects. The positive 
one is the "cooperative editing" effect: The article gets constantly 
improved and added upon, with many small parts leading to one great article. 
The negative one is that the first version may have been written by someone 
who does not know the subject very well, or is not a good writer. This may 
lead to a poor structure of the text, which is not easily improved upon. 
Also, when the improvement is better done by rewriting than changing the 
existing material, those who know more about the subject might go write 
their own piece on more specialist subjects instead. Perhaps it would be 
interesting to compare Wikipedia's notes to the edit history of the pages.

Andre Engels