Sunday, April 01, 2007

In Defense of the Wikipedia

By: Michael Akerman

Right, so it's been a rough semester, but that's no excuse for the blog still wishing everyone a happy New Year, though the sentiment still stands. It's just that the year is not so new anymore. In fact, it's acquired a dull sheen of pollen, I think.

So, a short post to swing into the back of things.

It's a Wiki'd World

As my years in college progress, I am expected to write more papers, essays, and articles for my classes. This is to prepare me for a dismal life of doing the same as I report my findings to my boss or to the scientific community (in real life, it doesn't look like it's so bad. Presumably, you're writing about something interesting). Inevitably, the papers require sources to prove both that you researched something and you're not blatantly plagiarizing the work of others. Nearly as inevitably, a major restriction is hung like an albatross around the assignment: no Wikipedia.

In the scientific world, sources are expected to be trustworthy (well, let's make that "in the world world"). For scientists, this means at least one or two scientists have taken a gander at the paper and found it logically and experimentally sound. This is a wise course to take, of course, because there is a surprising amount of junk science that tries to float out, and it often does even with these safeguards. To the majority of the scientific community, this restriction means that the scientific literature and traditional encyclopedias are fine. Wikipedia is not. The general reason for this is "anyone can write anything they want to!"

This statement is, of course, only slightly true. You can write anything you want to. You cannot get it to stay, and you cannot continue to write anything you want to unless it's true. You get IP banned quickly for willfully vandalizing the Wikipedia, and vandalism is found quickly. Whenever a page is updated, it goes onto the list of recently updated pages. With thousands of nerds reading that page at any time (they have nothing better to do), suspect information is quickly noted, and the page is marked to note that the information might not be true. If someone actually knows information counter to what is posted, it is changed as soon as it is seen.

I remind you that there is a large number of people checking the "recently updated" page, and that these are largely very smart people (they're reading an encyclopedia for fun). It's not hard for them to catch inaccurate information. Clearly, one cannot rely on their steadfast observance in a very obscure subject where an expert is unlikely to be reading, but Wikipedia is generally a terrible source of information for truly obscure subjects, so you wouldn't use it for that anyway.

I ramble a bit. My point is, the Wikipedia isn't a lawless project where people shoot from the hip and onto the internet. It is, in fact, the most peer reviewed source available for most subjects. In the traditional literature, you must rely on the expertise of a small group of people who are tasked with finding mistakes in reports and articles that are often only tangentially related to their field. On Wikipedia, there are enough reviewers that one of them, at least, will be more closely associated with the field in an article than any of the members of the peer-reviewing committees can boast.

This doesn't mean you should take the information on Wikipedia at face value; far from it. You shouldn't take anything at face value, from a newspaper article to a table in "Perry's Chemical Engineer's Handbook." The only way to be certain you get accurate information is to understand the subject and think critically. If information doesn't make sense, it's probably wrong. This applies to traditional literature and the information sources of the digital age. Exempla gratia, I'm working on a computer model for work that uses equations and data from a peer-reviewed article that has been around for some time. One of the graphs in the article is supposedly based on 150 g/L of starch mixture. After the reaction, the graph shows something along the lines of 200 g/L of a mixture of the starches. 50 grams of starch prestidigitate into the data. Something is clearly incorrect in this peer-reviewed paper from a widely-respected source (I'm fairly certain they actually started with 200 g/L, and wrote the wrong trial's data in the text).

The point is, you can't really just trust anything. You have to look at it with a bit of the cynic in you, at least when the accuracy of the information is really important. Wikipedia is no different than the traditional literature, and, for non-obscure subjects, is more reliable.

I think the scientific community will come to accept sources like Wikipedia, but the old guard is the one standing at the door currently. All they've ever known are the journals from organizations such as the American Chemical Society. They are the standards they've relied on, and they have a hard time trusting new things. I think it's the same thing that causes so many senior citizens to resist embracing emails and IMs. It's a harmless change that they are simply not used to.

By my hand,

~Michael Akerman


Michael J. Smith said...

Yeah, my professors do the same thing. In fact, there was study done, I can't remember by who now, of Wikipedia and Encylopedia Britannica entries which found that the two were virtually identical in accuracy.

As regards the self-correcting of Wikipedia, after Pinochet died earlier this year, I checked out his Wikipedia entry to find out that he was burning in the flames of Hell for the massacres he'd perpetrated against the Chilean people. When I refreshed the page, the attack was completely gone, the abuser's IP presumably banned. I still like Colbert's definition of Wikiality though: truth arrived at by consensus.