On Wikipedia
Stephen Colbert’s piece on Wikipedia has been getting a lot of coverage, and it got me thinking about the nature of factual errors in reference material.
The common wisdom is that sources such as Encyclopœdia Britannica are more accurate than Wikipedia. I am not aware of any actual studies which provide evidence as to the accuracy of this. The reasoning is that because anyone can edit Wikipedia, anyone can add incorrect information.
Which is true, but the question is not can people do it, but how much are they? And, again, I don’t believe there’s any hard numbers for that. There’s also the question of valid inaccuracies versus invalid. For example, if an effective cure for cancer was discovered, large amounts of any article referencing cancer would have to be edited to reflect this. Their information would be accurate, but out of date.
Which brings me to the point of all this: the time an error is in the wild is more important than the severity of the error.
If an error gets published in Encyclopœdia Britannica, it will remain there until a new edition is published. The current edition of the encyclopedia was first published in 1985, with revisions happening around once every two years. Additionally, the new printings are expensive; many people will keep the set they have for years, making the total lifetime of an error extremely high.
On the other hand, errors in Wikipedia are corrected extremely quickly. The cost of publishing is essentially zero, and so it the cost of consuming. So, even if you accept the premise that Wikipedia is less accurate, I feel that it’s comparable on an overall basis.
The real moral here is that you can’t believe everything you read, and you should have at least two sources when citing something. That goes for printed media as well as online sources.