Slashdot Mirror
Nitpicking Wikipedia's Vulnerabilities
tiltowait writes "A lot of Wikipedia critics point to hypothetical situations when giving reasons for not valuing the site. Wikipedia even has a 'Replies to common objections' article set up to field these. I'd rather look at some real examples of applying the same level of scrutiny to materials often held up as the Platonic ideal of 'scholarship,' such as peer-reviewed journals, conference papers, established journalism sources, monographs, and print encyclopedias. Even these have disclaimers because they can be can be vandalized or have their reliability and accuracy questioned. As dangerous as it is to trust unverified information, it can be just as bad to make prior judgments discounting information because the source happens to be anonymous. The above examples illustrate that all materials existing along a continuum of valuable information formats. Wikipedia articles can be useful for quickly obtaining factual overviews or as a starting point to further research. But that's just one librarian's opinion. How do tech-savvy people view Wikipedia?"
As dangerous as it is to trust unverified information, it can be even more dangerous to trust information which has been "verified" by "experts" (especially if it's information from your 1966 set of EB's)
Sure, Wikipedia probably contains more errors than EB, but it also contains many more articles. It would be interesting to know how these ratios compare.
So that's basically it, there is a spectrum of categories from where Wikipedia works well and has reliable information (mathematics, history and technology categories) to where it is just edit wars that get worse and worse (society and history categories). Wikipedia is fairly reliable about what ideas Godel had about mathematics, Wikipedia is completely unreliable if you are interested in reading about say France's Front National or Vietnam's National Liberation Front. Wikipedia has not gotten better over the years in this regard, it has gotten worse. There are left wing wiki encyclopedias like Demopedia, Dkosopedia and Anarchopedia, and right-leaning ones like Wikinfo, and I predict over the coming years these alternative wikis will become quite large.
One recent example I can give, one guy just popped up who is accusing virtually every left-wing or liberal person in the 1950's was a Soviet spy, and by virtually everyone I mean editing hundreds of biographies and inserting that they were spies. Doing this is fine if done in the right way, but he is a bit nutty or stubborn or whatever and he has a dozen people reverting his stuff but that doesn't do much good. Then we have Lyndon Larouche followers come in as well. Or way out communists saying nutty things. Wikipedia would probably be better off if these people all went off to their own respective wikis.
sure it couldn't be cited because the information there is simply too fluid and couldn't be counted on to remain unchanged over time.
If you're allowed to cite any other web page, why can't you cite a Wikipedia article. As long as you put the date you accessed it in the citation, what information was on the page is even less ambiguous than the webpage.
Ewige Blumenkraft.
A few days back there was talk about the Moller Sky Car, and someone said that the Newtonian and Bernoulli theories are incompatible, citing a Wikipedia article. (I'd link it, but I have a freakin migraine and really need to get to bed...)
Well, the wikipedia article was BS. Pulling out a real text like "Fundamentals of Aerodynamics" by Anderson would confirm that the Newtonian and Bernoulli views are compatible, just two different ways of expressing the same phenomenon. But since anyone who thinks s/he knows something about something can edit a wikipedia entry we get entries like that, which spread falsehoods.
I personally avoid Wikipedia for that very reason.
I suggest to people that when they are interested in a phenomenon that they try to find a reputable website that focuses on **just** that phenomenon. For example, if you have a question in aerodynamics, look for an aero website. Et Cetera.
-everphilski-
(Dislaimer - I'm a wikipedia administrator, arbitrator, and the "featured article director" -- I choose the featured articles you see on the main page every day)
/Lexika: Wikipedia gegen Brockhaus und Encarta/, starting on p. 132 - http://meta.wikimedia.org/wiki/Wikipedia_vs_Brockh aus_and_Encarta
Last week I was a guest speaker for a group of education graduate students about Wikipedia (the course was on technology use in education; wikipedia was part of the curriculum). Before the lecture, sent them a few items I thought they should read - objective studies of Wikipedia's accuracy done by impartial, outside organization. Here's what I sent them:
----------
1) "A group of students in the Graduate School of Library and Information Science at the University of Illinois has published a paper entitled "Information Quality Discussions in Wikipedia" (PDF format). The focus of the paper was on assessing the IQ of Wikipedia featured articles -- in this case, IQ stands for "information quality" -- when compared to other samples from the project, including featured article removal candidates, pages marked as NPOV disputes, and a selection of random pages. According to the paper, the study showed how seriously the Wikipedia project views issues of article quality. The authors concluded that as a quality standard, the featured article process "is not ideal, but it does seem relatively rigorous." They also noted that the process is not as resource-intensive as other possibilities, such as blind judging." - http://en.wikipedia.org/wiki/Wikipedia:Wikipedia_S ignpost/2005-08-01/Featured_content
PDF of research paper can be found at: http://www.isrl.uiuc.edu/~stvilia/papers/qualWiki. pdf
2) An article comparing the WP to Brockhaus and Encarta has appeared in issue 21/04 of C't, a major German computer engineering magazine. It is titled
Full survey results can be found at: http://mail.wikimedia.org/pipermail/wikipedia-l/20 04-October/035339.html
3) "As publicly editable sites, Wikis are vulnerable to vandalism. We've examined many pages on Wikipedia that treat controversial topics, and have discovered that most have, in fact, been vandalized at some point in their history. But we've also found that vandalism is usually repaired extremely quickly--so quickly that most users will never see its effects." - IBM study of Wikipedia - http://researchweb.watson.ibm.com/history/results. htm
4) Computer Science professor (and minor geek rockstar) Ed Felton (http://en.wikipedia.org/wiki/Edward_Felten) posted in his blog about a
small-scale survey he did of Wikipedia: http://www.freedom-to-tinker.com/?p=674
-----------------
As far as my personal interactions - as featured article director, I can say first-hand that we've been hitting really hard on the need to have inline cited sources in the article text. It's been an explicit requirement for featured articles for some time now (9-12 months or so). In many ways, this makes our content much more trustworhty than most other information sources.
Furthermore, purely from personal experience, I can say there's something to be said for the expert-hobbyist. For example, the "best" writer on wikipedia (in terms of number of featured articles written) is a 17 year old from New Jersey who writes long, thorough, well referenced, accurate articles on, erm, British and the Bri
To make laws that man cannot, and will not obey, serves to bring all law into contempt.
--E.C. Stanton
Not to nitpick, but NPOV is only one of three supreme rules, the other two being No original research and Verifiability. They must be all taken together, no one trumps the other.
So, if you're an expert, add a "Sources" section with references to back you up, that prove that your statements aren't just things that you believe, but are indeed the consensus of the experts in your field. "No original research" means that, in fact, Wikipedians explicitely are NOT equiped to judge whether something is an expert consensus or not. So, as long as you back your statement up with published sources (just as you'd do in an academic paper), you should be fine.
Also, if someone reverts you, and you know you're right, don't back down. Everyone on Wikipedia needs to be both bold and civil. Go to the "discussion" part of the article, explain that you're an expert, explain that you think this is also the consensus of other experts, and if they're being civil, they'll welcome your edits with open arms.
This has a rather painfully simple technical solution. Just demand that your students cite specific version of an article by URL. For example, the *current* article on Slashdot at Wikipedia is:
t &oldid=24939679
http://en.wikipedia.org/w/index.php?title=Slashdo
Someone will probably modify the Slashdot article itself in the next five minutes. But the URL I give will retain unchanged forever. Of course, if the edit someone does in five minutes improves the information, my snapshot URL won't present the improvement. But then the future improvement isn't what your students read either.
Buy Text Processing in Python