« Trade-offs | Main | Very good sentences »
Cooked books
If I had to guess whether Wikipedia or the median refereed journal article on economics was more likely to be true, after a not so long think I would opt for Wikipedia. This comparison should give us pause.
That's me, writing for The New Republic. But what does this all mean? ("Sadly, the final lessons here are brutal.") I consider the recent spate of fake biographies and memoirs and arrive at some conservative and traditionalist answers.
Posted by Tyler Cowen on March 13, 2008 at 08:11 AM in Education | Permalink
Comments
Good for you. I'm sick of all of the wikipedia bashing. Let's recognize it for what it really is; one of the single greatest achievements in human history.
I work in a public middle school, where many teachers tell students not to use wikipedia. This is absolutely crazy. Not only does it essentially prevent students from doing any research at all on their own; but wikipedia has far more detailed and accurate information than their textbook; "Roosevelt's New Deal pulled America out of the Great Depression by creating jobs for the unemployed."
Posted by: josh at Mar 13, 2008 8:38:43 AM
In my experience, wikipedia is often very good at presenting many dissenting views, especially if one follows the references. I'd say this gives readers a much healthier view of the 'truth' than a small collecting of authors in an economics journal, although I admit to never having read an entire journal cover-to-cover.
Posted by: Grant at Mar 13, 2008 8:40:45 AM
By the way, I'd really like MR to give its opinion on Google's upcoming wikipedia competitor: Knols. My thoughts on it are that Knols are more likely to represent popular knowledge, but less likely to express the diversity of opinions that Wikipedia does.
I think its also worth mentioning that the Wikipedia project was supposedly inspired by Hayek's The Use of Knowledge in Society.
Josh, don't you think the anti-wikipedia bias might stem from a simple fear of competition? Teachers and professors of all sorts don't want to be replaced. I know I've used Wikipedia in preference to textbooks and the notes of professors in graduate engineering courses; I'd imagine a few people feel threatened by this.
Posted by: Grant at Mar 13, 2008 8:48:29 AM
Not an apt comparison. Journal articles attempt to do what wikipedia does not: present original research, which by its nature is far more likely to be wrong.
The correct comparison is between wikipedia and the median undergraduate economics textbook.
Posted by: Cyrus at Mar 13, 2008 8:49:54 AM
Grant,
Probably a combination of that, and just plain stubborness. Most adults are not very internet savvy; at a department meeting I mentioned imdb, and not a single other person had heard of it. These teachers just get it in their heads that anybody can edit wikipedia, and assume that means it is not reliable and thus not useful or counterproductive. QED. If only they knew how unreliable the things they teach students are.
Posted by: josh at Mar 13, 2008 9:10:06 AM
I'd say that one of the strongest sides of Wikipedia is rarely even mentioned: the links sections below the articles. In many cases, I found them extremely useful as a shortcut to the information on the subject available on the rest of the Internet that can save lots of tedious googling. Even if Wikipedia articles contained nothing except these links sections, they would still be worth checking out when researching any topic except perhaps one's narrow field of expertise.
I've also notice that in the last year or two, most Wikipedia articles have immensely improved in that they now provide external references for any extraordinary claims. Even if they don't, they are now usually marked with clear caveats about missing references.
Posted by: Ivan at Mar 13, 2008 9:57:22 AM
Cyrus is right. Wikipedia explicitly forbids original research. Journals explicitly require original research -- they won't publish stuff that's already well-known and widely proven.
To the extent there is a difference in reliability, it probably has a lot more to do with the difference in types of articles than the differences in the reviewing or publishing process.
Posted by: A student of economics at Mar 13, 2008 10:40:30 AM
I always told my students that I never, ever wanted to see Wikipedia cited in their bibliographies, but I certainly directed them there as a good starting point, and a way to get a quick and easy broad overview of whatever issue was at hand. Wikipedia is great, but a user has to be aware of its limitations.
Posted by: bartman at Mar 13, 2008 11:19:26 AM
Sorry Tyler, but I am not convinced.
The median journal article focuses on what we don't know and what is not so easily verifiable. The median Wiki entry focuses on what is well established (but not known to all). Wouldn't one expect (and even demand) more errors in a activity which is meant to create new knowledge (such as academic research) than one would in a synthesis of existing knowledge.
Consumers of economic knowledge need to be forewarned, however, that journals are not repositories of facts, but rather of ideas.
Roger
Posted by: RogerClemens at Mar 13, 2008 11:48:43 AM
Tyler,
In an undergrad class would you accept wikipedia as a source in a paper? I don't for my econ development class. Like previous posters I direct students there and often use wikipedia to look up econ concepts. By not allowing students to use Wiki, I think it gets them to try to look for other sources.
Posted by: Seth at Mar 13, 2008 12:17:22 PM
Regarding use by students, I agree with bartman and Seth. I suggest that students use Wikipedia
as a good initial source, but that they then should track down more reliable basic sources to be
cited in papers.
I have read all the claims of how Wikipedia is at least as reliable as the Encyclopedia Britannica,
which may well be true. However, I must report that I have found a number of egregious errors in
Wikipedia entries. It may still be that there are fewer than in journal articles, but then it has
already been pointed out by others that the two are not comparable at all.
Posted by: Barkley Rosser at Mar 13, 2008 1:03:18 PM
Tyler,
I am interested to know if this also applies to other instances where large groups of people working independently achieve equal or better results than a small group of experts or professionals (if this is indeed the direction you were headed).
Is this a ubiquitous economic (and/or social) law or does it succeed sporadically?
Posted by: Ocean at Mar 13, 2008 2:38:26 PM
Free Exchange at the Economist picked this up.
Posted by: Dave at Mar 13, 2008 2:58:12 PM
Ocean, the short answer is (I think) "yes". Google prediction markets, and incentive-compatible mechanisms which give people incentives to reveal private information.
Posted by: Grant at Mar 13, 2008 3:28:05 PM
Prediction markets, the hobbyhorse of Robin Hanson and overcoming bias,
are good at some things (predicting election outcomes) and very bad at
some others, e.g. forecasting Nobel Prize recipients.
Posted by: Barkley Rosser at Mar 13, 2008 3:48:19 PM
Ouch! The economist downgraded you to Mr. Cowen. That's Prof. Cowen or Dr. Cowen to you, peeps.
Posted by: J. at Mar 13, 2008 3:49:18 PM
Picking nits, the insightful review on Amazon of Defonseca's book was not the first Amazon review. Amazon now moves the most helpful reviews to the first page for the book.
Posted by: Lemmy Caution at Mar 13, 2008 6:11:16 PM
Picking nits, the insightful review on Amazon of Defonseca's book was not the first Amazon review. Amazon now moves the most helpful reviews to the first page for the book.
Posted by: Lemmy Caution at Mar 13, 2008 6:12:41 PM
I recently took over an engineering project that a previous student had used to complete her master's degree, and my first task was to read her thesis. To my shock, it cited Wikipedia, not once, but twice!
My thoughts on citing wikipedia is: why bother? Isn't the rule of thumb is that you don't have to provide a reference for knowledge that can be found in multiple sources? Isn't that pretty much the requirement for a wikipedia article?
Posted by: Yogi at Mar 14, 2008 4:08:46 AM
Citing wikipedia for a high school level class is certainly no different than citing a text book. For undergraduates, it seems reasonable that they cite wikipedia, though if you expect some level of external research for the paper, you may not regard wikipedia as much of an effort in that regard, but it certainly doesn't detract from the effort. The thing to keep in mind that in high school and for undergraduates, citing sources is just practice for what they might do as graduate students, or just proof that they actually did some external research.
With an actual research paper, citations are used to validate claims and wikipedia may not be that useful in that regard. Obviously, they don't include original research. And as secondary sources, they are anonymous, so their credibility is in question. Of course, for non-controversial claims maybe this may be fine. Wikipedia will provide mathematical arguments, however. This is kind of like original research in that they explain how a result is reached from more basic principles. Thus, I can see a valid basis for a wikipedia cite in an engineering type document.
Posted by: mpowell at Mar 14, 2008 2:43:57 PM
"don't you think the anti-wikipedia bias might stem from a simple fear of competition? Teachers and professors of all sorts don't want to be replaced. "
My experience, not a single professor has ever mentioned this in the least!
Its really very different functions indeed..
Posted by: edwardseco at Mar 15, 2008 10:34:34 PM
