The perils of Wikipedia

In an entry below I offer that the citation of a Wikipedia reference is not reliable, and I can't take responsibility if someone changes the entry between my link and your click. I am not totally kidding, I "Wikipediaed" a semi-famous individual recently and the entry described him as a serial rapist. In broken English. Someone was obviously bored, or had a bad experience with this small time celeb. I quickly reedited it, but it sure brought home to me the problem with Wikipedia. But then I thought: could you, as a blogger, just reedit or write your own Wikipedia entries and then link to them as a citation?

Anyway, I guess you could say I'm pretty cautious about linking to Wikipedia in regards to things that I'm not totally sure about, or that are not copiously cited. A lot of the technical stuff is good, and Wikipedia is actually often well referenced on many entries, so it is a good starting point, but it sure isn't The Answer (and remember, Iverson's field goal % isn't that high). There are similar problems with google. For example,

a) You have hypothesis/assertion A

b) Someone calls you on it

c) You proceed to use google (and google scholar) to find 10 citations which back you up

d) But there is a problem, the set of all studies relating to hypothesis/assertion A is enormous, it is very easy to simply return back an unrepresentative sample of studies which do support your particular hypothesis via a query

e) So, two things, the citations need to be unbiased and representative of the totality of the scholarship. And second, you need to actually know something about the field or the topic so you can use google intelligently. Most people who I have had this issue with are offering up an unrepresentative sample without knowing because they don't know the shape of the distribution a priori.

Update: From the comments:

A citation or link to Wikipedia should be in citation format -- which means its unchangeable.

For instance, http://en.wikipedia.org/wiki/Rowan_Oak takes you to the ever changing (or not) description of Faulkner's home.

Conversely, http://en.wikipedia.org/w/index.php?title=Rowan_Oak&oldid=71987882 takes you to a time/date frozen version which will never change.

To get the citation link, click on the "cite this article" link under the search box when viewing the main article.

Tags

More like this

Hard on the heels of Wegman's farcical attempt to sue Mashey comes Watts's incompetent attempt to meat-puppet wiki. If you want to see my comments at WUWT that didn't survive moderation, you'll need to read stoat spam or just imagine them; I said nothing that wasn't obvious. My favourite, I think…
Interesting conversation at lunch today: topic was academic performance metrics and of course the dreaded citation index came up, with all its variants, flaws and systematics. However, my attention was drawn to a citation metric which, on brief analysis, and testing, seems to be annoyingly reliable…
The October issue of the Journal of Biological Rhythms came in late last week - the only scientific journal I get in hard-copy these days. Along with several other interesting articles, one that immediately drew my attention was Clock Gene Wikis Available: Join the 'Long Tail' by John B. Hogenesch…
"Part I" is very presumptuous. I might never write part II. Ah well, I press onwards in hope. I'm going to take my text from Climategate: the corruption of Wikipedia and see what we can learn about wiki's workings from the way people misunderstand it. I should warn you that blog is mostly recycled…

When you say "Wiki", do you mean "Wikipedia"? The two terms aren't equivalent. A "Wiki" is a general term for a site that runs "wiki" software-- which maintains wide collaboration, user-editable web pages-- of which there are several different implementations. MediaWiki is the software used by Wikipedia. Wikipedia is the specific internet encyclopedia site which uses a Wiki to manage its content.

I'm guessing from context that when you say "Wiki", you mean "Wikipedia", but it's not 100% clear. What you discuss as the issue for Wikipedia would of course apply to many Wikis-- any Wiki that doesn't have a restrictive registration process for editing content.

-Rob

A citation or link to Wikipedia should be in citation format -- which means its unchangeable.

For instance, http://en.wikipedia.org/wiki/Rowan_Oak takes you to the ever changing (or not) description of Faulkner's home.

Conversely, http://en.wikipedia.org/w/index.php?title=Rowan_Oak&oldid=71987882 takes you to a time/date frozen version which will never change.

To get the citation link, click on the "cite this article" link under the search box when viewing the main article.

For me Wiki plus Google are invaluable as my first source, and I often rely on them for non-controversial questions of secondary interest to me. (Not things I know nothing about, and not things I know a lot about.) You do need a bullshit detector and enough knowledge about the field to read critically.

All recial, national, political, and religious questions are controversial, of course.

By John Emerson (not verified) on 26 Aug 2006 #permalink

The best advice about wikipedia is to treat it like a knowledgable friend. It's not an authoritative source, but it's a pretty good place to start many times.

I'd agree wih Chris and John. I find Wikipedia quite useful as a first initial stop, and I often link to it on my own blog for quick and easy definitions, etc. But I woudn't use it as a primary source/citation for anything more formal, like an article, a paper, or a book -- not without verifying the information there in at least 1 or 2 other less, shall we say, "fluid" sources. :)

Chris sez: "The best advice about wikipedia is to treat it like a knowledgable friend."

Ya, one that gets drunk time to time and does hard drugs.

The Wikipedia has some serious errors occassionally. I once found a "Thanksgiving Day" article that started "Thanksgiving, also known as Nigger Day..." I reported the error immediately. Wikipedia articles are generally edited a jillion times by a small army of people. They are much better than blogger articles or etc as reference points, because they are objects of mass editing.

I have a major problem with Wikipedia bashing, because academics and journalists make errors too - sometimes, small ones that nobody notices. And sometimes big errors (whether goofs or more agenda-determined Big Lies) are made that persist for years or generations. And unlike Wikipedia (which tends to have stuff written on even arcane topics), academia and journalism are often guilty of major sins of info /omission/.

"Reader beware" applies to Wikipedia just as everywhere else.

"Reader beware" applies to Wikipedia just as everywhere else.

Indeed.

For a casual reference, Wikipedia is great.

For a lot of things, Wikipedia is better than a "traditional" encyclopedia, because (a) it's more up to date, and (b) traditional encyclopedias fail to cover in minute detail the stories of very minor characters in video games.

And, of course, we all learned in something like 5th grade that we can't use an encyclopedia as a primary reference.... Indeed, when I was in fifth grade, there was no World Wide Web (though there was an Internet, but I knew nothing of it), never mind Wikipedia.

-Rob

One other thing on web citations -- I've seen some journals and such that state you should give a date when the site was accessed. With something like Wikipedia, that keeps an edit history, one could in principle reconstruct what was read. With other sites, it's harder, although you might get an approximation with the Internet archive.

The transient nature of the web is why we will always need "Journals." They need not always be in print, but there will need to be "releases" that have definitive versions of things. (Much as the Linux kernel, despite being a big collaborative project, has specific releases that signify a unique, well-defined set of code.)

-Rob

Above I lumped Google and Wikipedia. Actually, as more stuff is posted, Google gets better and better, and often can take you to exactly the thing you need, quicker than a library and just as soundly.

For things of peripheral, casual interest the net is utterly fantastic. You can quickly get a lot of information about very obscure topics. There are usually big holes, but the quickness is fantastic.

Wikipedia has a further problem in two areas.

First is the policy of keeping information value neutral. Which more often than not winds up raising worthless information to a status it does not deserve.

Second is the policy of presenting word definitions that do not, in any way, match how the word is being used currently, or has been used in the past.

As a lead to further information it is good. As a primary source, not so hot.