If the standard is which encyclopedia most resembles the Encyclopedia Britannica, then I’ll concede the Encyclopedia Britannica is clearly superior to Wikipedia.
Wikipedia should never be used as a sole or even main source of information. If he needs to research something that books are written about, then nothing beats a book. This is the conclusion I have come to after years of trying to get research information off the internet. If he can’t get books on the subject then he should look for websites devoted to it, not Wikipedia, though the external links are pretty useful.
Actually most of the articles from those really old encyclopedias are already in Wikipedia: we imported nearly the entire text of the (now public domain) 1911 edition of the Encyclopedia Brittanica some years back.
And if I catch you deliberately inserting false information into an article, rest assured that I will block you. 
I think I have said it before, but on behalf of all Wiki-Fans: Thank You and the rest of the Wiki Staff for giving us one of the best free tools ever.
Jim
Neither should any encyclopedia. If you are writing a paper for college, many teachers will tell you from the front not to use an encyclopedia entry as a cite.
When I was in high school, the use of an encyclopedia as a source was strictly forbidden. I would imagine the same would be true of college.
Encyclopedias are tertiary sources at best, and only slightly above water cooler talk in terms of reliability.
Are they really that laden w/ errors?!! 
NO WAY!
A tertiary source has been cooked twice already. The chance of the information you seek being degraded, either by mistake or through bias, is quite high. Ideally one would only use primary sources, but in many cases this is not possible (primary sources often being hard to obtain access to).
In the early 90s I was, for the hell of it, looking up Alan Turing at my local library. The article in the first encyclopedia did not mention how he died. I was curious and looked him up in every encyclopedia they had. There were many conflicting details in the articles, not just surrounding his death. This is a man less than a hundred years dead! Many times when I look up something I know well in an encyclopedia, I am appalled at the misleading statements in the articles. Many times, I can tell the kind of error the author made. Equating things that are not the same, over simplifying already simplified explanations, presenting theories as axioms or vice versa and most often, presenting commonly held beliefs as hard fact.
This issue is not unique to encyclopedias. One thing college research is taught me is not to underestimate the ignorance one individual can spread. When looking into a subject, trying to pinpoint where one belief or idea started, often many authors rely on secondary or tertiary sources and often all of what appear to be secondary sources are actually relying on a single primary source that on close reading is secondary and just presented as a first hand account summarized from other accounts and put into the first person.
Holy crap dude… who knows how many things we accept as fact were just typos or ULs when they got started?
(insert funny historical-themed joke here)
See, this is where mistakes start. Your remark seems to call me dude. Dude normally refers to those of the male sex. I am a natal woman. Yet, others could take your remark as corroborating evidence that either I am male, or that I asserted myself to be male. I have heard others take the description that a natal women was from the city of Natals, so it is even possible my correction could introduce further error in belief.
So did you know that Lee was a male nurse in a Neo-Natal Unit?
What I thought this was how it works?
You mean he’s a medic for the skinheads? I hate those fuckers.
Hey now, I have skin on my head.
Has anyone done any studies on datamining wikipedia? I imagine that of the 800,000 articles, it would be something like 5% of the articles constitute 95% of the page loads and 1% constitute 80% or something similar. Furthermore, has anyone done studies on the resistance to defacing? I know I’ve heard lots of anecdotal studies and stories but nothing solid. A defacing would be relatively easy to tell, it would involve an edit which involved large chunks of text being changed and then a reversion back to the previous version. I would also imagine that something like 95% of defacings happen on 0.1% (ie: George Bush) of the pages and that they are reverted relatively quickly. But also if one were interested, it would be easy to construct a orchestrated series of defacings that would be very hard to defend against.
If some studies have been done, I would love to take a look at them. Alternatively, if the data exists, I might consider doing a bit of mining on my own. KellyM, would you know someone who could get access to that data?
Even measuring page loads is difficult. Wikipedia runs in a clustered environment, meaning that any particular web request will be serviced by any of about 20 different servers (this number varies from day to day as servers go in and out of service). Each maintains its own log. In addition, all of the servers are behind front-end cache servers, which cache copies of commonly requested pages to reduce web engine workload. In order to gain an accurate representation of page load count, you’d have to aggregate the logs of all the Apache webservers AND all of the Squid front-end cache servers. We used to have a script to do this, but it is presently broken and nobody has bothered to fix it. As a result, we have no real idea what pages are most commonly visited. We do have total page view statistics; recently, we peaked at 6,000 requests per second (aggregated across all servers).
We detect most vandalism through sideview monitoring of the change log, rather than through randomly stumbling across it. Under normal circumstances nearly every anonymous edit gets looked at by at least one RC patroller. The edits would have to be subtle vandalism to get past the RC patrollers; furthermore, a user found vandalizing will have his recent edits checked for other vandalisms. Defacing events are more common on “high traffic” pages like George W. Bush and Pope Benedict and other such pages, but people also vandalize articles about obscure mathematical concepts. (Some people actually hit the random page button and then vandalize whatever page they get.) IBM did a study on vandalism a couple years back and determined that most vandalism is reverted within 5 minutes. In recent months, it’s my experience that most vandalism (at least by anonymous users) is reverted within one minute and often within seconds.