78

I ask this question because in school, students are always told "do not quote wikipedia" or that "Do not believe everything you see on Wikipedia, we are teachers and we know when something is wrong there". (The latter being more common in my case) Anyway, Wikipedia does have an article about it here. Still the question stands about whether it is reliable, and if it is, when or if it will be permissible in the classroom.

Have any studies been made on the reliability of Wikipedia?

Sklivvz
  • 78,578
  • 29
  • 321
  • 428
picakhu
  • 3,158
  • 1
  • 22
  • 42
  • 2
    Don't think this is a the right place for that question. I'll say that Wiki is ok, but mostly when I look up something in it, I go down and hit the references. – Mike Apr 10 '11 at 16:45
  • I've modified the question so that is not off-topic. – Sklivvz Apr 10 '11 at 18:07
  • 2
    Side question. Assume I accept that wikipedia is asymptotically correct, possibly even authoritative. (And that is, in fact how I treat it.) How do I know that this version of the page I have in front of me is good? Well, you check the edit history, and use look at the sources listed, of course. But if you are doing that, are you really treating the sources as reliable? – dmckee --- ex-moderator kitten Apr 10 '11 at 22:00
  • 6
    Just a comment. For subjects I am not familiar with, Wikipedia seems pretty good. For subjects I *do* know something about (such as software optimization), Wikipedia calcifies widely-held misconceptions. So take your pick. – Mike Dunlavey Apr 11 '11 at 00:52
  • 2
    I am of the opinion that you shouldn't quote *any* encyclopedia. The purpose of an encyclopedia is to provide an introductory amount of information about a topic in a manner that can be easily understood. If you're using an encyclopedia as a complete reference on a topic, you will be farther from the source of the information than if you use the encyclopedia to gather a list of other references that are more accurate. – zzzzBov Apr 11 '11 at 02:45
  • I think it is banned from the classroom because it is just too easy for a student to go and copy an answer straight from there. Need to teach some better research skills than that. – Craig Apr 11 '11 at 05:04
  • 2
    I personally use Wikipedia as a pointer. In my classes, I desire that students avoid citing Wikipedia due to the open editing policy; however, I do inform them that using Wikipedia to discover othe sources or related information is acceptable. –  Apr 11 '11 at 00:09
  • Response to the [edit]: If Wikipedia cites a source, the source stays the same. Therefore, analyse each source as if it were something you found on Google. – Dale Apr 11 '11 at 01:37
  • 1
    @Craig: perhaps that is right, but in any case, even professionals go to Wikipedia first. – picakhu Apr 11 '11 at 07:08
  • 3
    Let's keep in mind that paper does not convey any inherent benefit, but people still tend to have that prejudice. For example, if you reference a printed book as a source for your essay, your teacher will accept that without a question, but if you quote wikipedia they'll reject the reference. The prejudice is that the teacher has no idea about the validity of the printed words in the book, but gives them more weight just because they are on paper. However, lots of books get printed without fact checking by experts. Books don't get peer-reviewed, but wikipedia entries do. – Kenny Wyland Apr 11 '11 at 23:00
  • @Kenny Wyland: They're likely not to accept a paper encyclopedia either. There is a prejudice in favor of books (which is reasonable; nobody's going to change my copy of a book while I'm not looking), but that's not the only thing about Wikipedia. – David Thornley Apr 12 '11 at 12:32
  • 1
    @Mike Dunlavey: Your position on software optimization isn't mainstream, and I wouldn't expect to see it in an encyclopedia. For subjects I know something about, Wikipedia is usually excellent (unless there's political reasons why it isn't). – David Thornley Apr 12 '11 at 12:34
  • @David: On that topic, you might find this interesting: [Conservapedia](http://www.conservapedia.com/Main_Page) – Mike Dunlavey Apr 12 '11 at 14:38
  • 1
    @David Thornley, I'm not referring to just print encyclopedias. I'm talking about any printed material. For example, if I'm writing a paper about the War on Terror and I quote something from wikipedia the teacher will reject it. However, if I quote something from The Dark Side by Jane Mayer the teacher will accept it. Why? It's the paper prejudice. The teacher has no idea if Mayer's facts have been checked, but they are given more credence because they are on paper. – Kenny Wyland Apr 12 '11 at 15:23
  • @Kenny Wyland: An on-line encyclopedia and a dead-tree book aren't comparable. I know nothing of that book, but it's a source that isn't an encyclopedia, that was specifically written about a topic. What would the teacher think of using a well-written website about the War on Terror as a source? – David Thornley Apr 12 '11 at 16:40
  • 2
    @David Thornley, I can't speak for this particular teacher, but my experience with teachers in the past is that they will (strangely) accept a website source more readily than they will accept a Wikipedia source. The editable nature of wikipedia scares people, but without true reason. Some flat-out-lying wacko could have made the well-written website, but people seem to have a prejudice against wikipedia even though it's like to be more accurate than something posted on a blog due to having a community who checks the facts and corrects problems. – Kenny Wyland Apr 12 '11 at 19:56
  • @Kenny, you are indeed correct on this one, there is a ton of bullshit websites out there, some of which are "relativity (is) fraud...", "flat world" ... And if you are (very) careful on how you craft your argument, it will be accepted by a teacher.. – picakhu Apr 12 '11 at 21:49
  • @Kenny: There's nothing strange about it. Teachers typically don't want encyclopedias as sources. I agree that there are numerous screwy websites out there, and I have some seriously screwy books, but at least that represents research. – David Thornley Apr 13 '11 at 02:26
  • @David, I think the essential difference between what we are saying is that while you say that teachers don't _want_ encyclopedias as sources (because they want deeper research) I'm saying that teachers will _reject_ wikipedia. Despite their preference against encyclopedias, teachers will allow a print reference and not a wikipedia reference. – Kenny Wyland Apr 13 '11 at 03:51
  • @Sklivvz, not many people (almost none) are replying to the addendum question, about whether going to Wiki for the sources is reliable. (maybe wiki has some bias that I am not aware of). Should that be a new question or can something be done about it? – picakhu Apr 13 '11 at 05:51
  • @pica, removed the addendum - it should be a separate question. – Sklivvz Apr 13 '11 at 08:40
  • 4
    I assume you are talking about English Wikipedia, but it would be nice to state that explicitly. Because many users may assume that what is true for English Wikipedia is true for every other Wikipedia. I.e. if en.wikipedia is reasonably accurate, so is every other wikipedia. In practice they are unrelated as far as info and even rules go, and the quality varies wildly. (P.S. I am talking "legit" Wikipedia in different languages, not about conservapedias or other bs). –  Apr 13 '11 at 08:59
  • 2
    Wikipedia itself acknowledges that it's possible to [abuse Wikipedia guidelines and policies](http://en.wikipedia.org/wiki/Wikipedia:Gaming_the_system) in order to e.g. push subjective views into articles. This has happened before and will happen again. – StackExchange saddens dancek Apr 15 '11 at 11:48
  • 1
    Wikipedia's openness alone should preclude its use as a source for citations. Even if we can demonstrate that a randomly sampled Wikipedia article were 99.99% likely to be accurate, that doesn't prevent the user from gaming the situation by creating a false identity and manipulating (or all-out fabricating) the specific article s/he would like to cite. – Faust Apr 15 '11 at 20:48
  • Reliable implies that you get consistent answers from one day to the next. This is not the case with Wikipedia. An answer can be deemed correct on July 4th, and then incorrect on December 24th, due to editing. An Encyclopedia is printed once (for the most part), and the answer you see in print is the same answer you see the next day. Thus, Wiki may be just as accurate, but not just as "Reliable" – Ralph Winters Apr 11 '11 at 18:02
  • 2
    It's as reliable as Stack Exchange ;) – The Student Dec 06 '11 at 12:32
  • Scientific papers are not always reliable too. I do some paper reviewing for journals in computer science and I can tell you that many reviewers don't do their job correctly. So the question is, which sources are really reliable? None. The best is to find more than one. – Zonata Sep 02 '12 at 18:50
  • Here is a reliability order for questions that belong to mathematics: SIAM / AMS Journals > http://arxiv.org/ > Books published by "Springer Vieweg Verlag" > Wikipedia > .edu pages > your math teacher. Note that for single statements Wikpedia can be as good as everything that is cited when you check the cite. – Martin Thoma Apr 06 '14 at 18:40

5 Answers5

55

In December 2005, the science magazine Nature conducted a study to determine whether Wikipedia was as accurate as traditional encyclopedias, namely the Encyclopedia Britannica.

In Study: Wikipedia as accurate as Britannica, Daniel Terdiman summarizes the study results:

Wikipedia is about as good a source of accurate information as Britannica, the venerable standard-bearer of facts about the world around us, according to a study published this week in the journal Nature. ...

For its study, Nature chose articles from both sites in a wide range of topics and sent them to what it called "relevant" field experts for peer review. The experts then compared the competing articles--one from each site on a given topic--side by side, but were not told which article came from which site. Nature got back 42 usable reviews from its field of experts.

In the end, the journal found just eight serious errors, such as general misunderstandings of vital concepts, in the articles. Of those, four came from each site. They did, however, discover a series of factual errors, omissions or misleading statements. All told, Wikipedia had 162 such problems, while Britannica had 123. That averages out to 2.92 mistakes per article for Britannica and 3.86 for Wikipedia.

Of course, what makes Wikipedia different from a standard encyclopedia is that it can be updated by anyone and those updates appear immediately. When there's a particular hot topic item in the news, it's not uncommon for its Wikipedia entry to be updated rapidly, both by level-headed and factual participants and by more extreme, agenda-driven actors on both sides of the issue at hand. Consequently, when you read an entry at Wikipedia it's important to bear in mind that what you are reading now might include some dogmatic if not outright incorrect statements.

As Bibhas noted, it's important to check the sources and references. I also would encourage you to note the history of the piece you are reading. If there were many edits in a short window of time that may be a sign that there is some back and forth going between two "sides," which could imply that the information presented might not be as objective as it was prior to the volley of edits.

Community
  • 1
Scott Mitchell
  • 5,923
  • 4
  • 35
  • 32
  • 43
    I dislike it when people cite the Nature study. It puts factual errors and omissions at the same level. Most errors on Wikipedia were factual errors; most errors on Britannica were omissions, things that the experts feel should have been mentioned. Factual errors are obviously graver mistakes, since it is misinformation. – Borror0 Apr 11 '11 at 00:43
  • 30
    Wikipedia has changed a great deal in the last 6 years... – zzzzBov Apr 11 '11 at 02:43
  • 11
    @Borror0 [Quantity always trumps quality](http://www.codinghorror.com/blog/2008/08/quantity-always-trumps-quality.html), where Wikipedia is Quantity, and Britannica is *sometimes* Quality. While you shouldn't rely on Wikipedia wholly for a critical part of your Nuclear Warhead project you're working on in your spare time, for 99.9% of cases, using Wikipedia is OK. – Mateen Ulhaq Apr 11 '11 at 03:28
  • 7
    Note that the study does not exactly cover what the OP is asking for. *Reliable* is not the same thing as *accurate*. I would say that neither Encyclopaedia Britannica nor Wikipedia are reliable sources, although they are somewhat accurate. – Sklivvz Apr 11 '11 at 05:51
  • 12
    The problem with this analysis is the Wikipedia has a long tail of articles that very few people care about. These articles are not very reliable – Casebash Apr 11 '11 at 12:11
  • 1
    They ought to do a more modern analysis. However, in Wikipedia's defense, omissions can sometimes be more of a problem than errors, because they can be "unknown unknowns". – Andrew Grimm Apr 11 '11 at 12:16
  • 1
    From the supplemental materials: **Q:** Were the reviewers told that the articles were either from Encyclopaedia Britannica or Wikipedia? **A:** They were told that the two entries were from Britannica and Wikipedia, but they did not know which was which. Entries were stripped of metadata and formatting to leave only raw text, to preserve the blind as much as possible. Not from the SM: **Q:** Did the reviewers have internet access? **A:** Da-dung. – Ruben May 30 '11 at 21:59
  • A german study from 2007 showed that Wikipedia was better in being up-to-date and in the correctness than the German "Brockhaus Enzyklopädie". Here is the article: http://www.stern.de/digital/online/stern-test-wikipedia-schlaegt-brockhaus-604423.html?nv=sml – Martin Thoma Aug 28 '12 at 16:13
  • 2
    So, wikipedia says wikipedia is reliable. Does anybody else have a problem with that as evidence? – n00b Jan 25 '13 at 23:56
  • A lot of technnical articles have large inaccuracies. I recently was reading some articles about Unix in which **all** of the quoted sources, did not actually mention or even imply the fact that they were linked to in the article. A lot of the algorithms in Wikipedia as well have minor inaccuracies. I had a really painful experience trying to implement the A* algorithm by referring to Wikipedia. – asheeshr Mar 06 '13 at 02:40
  • The problem is that you cannot compare it to the Britannica by the number of factual mistakes. Wikipedia articles can be misleading and politically and ideologically loaded without containing factual mistakes. Wikipedia contains a large amount of "media coverage" of events, listing the opinions of various sources. These opinions can be vastly out of proportion. For example, one can easily "defend" an idea on Wikipedia, in citing that "neo-nazis criticized it...". It happens way too often. Just count the weasel words (some claim, studies prove, etc.) in Wikipedia, and compare it to Britannica. – vsz Aug 09 '13 at 12:46
19

The source of the data in Wikipedia is people like us, who knows about a certain topic or domain. And the sources of those data are shared at the bottom of most article. (For example, http://en.wikipedia.org/wiki/Horse#Sources.) You can check that on every single page as it has a discussion section where questions against the data, if any, are raised and proofs are provided if necessary (http://en.wikipedia.org/wiki/Talk:Pearl_Harbor). You yourself can raise question if you want to, they'll answer.

Hence, quoting Wikipedia is far more reliable than quoting a teacher him/herself. As the data is available there in each and every article, it has been filtered for years by millions of teachers like the one who teaches you.

If your teacher still don't agree, go to Wikipedia, check the sources of their data at the bottom of the page and provide that to the teacher.

P.S. - If sources are not available, check the references. They are always there...

Peter Mortensen
  • 190
  • 7
Bibhas
  • 315
  • 1
  • 4
  • 9
    Welcome to Skeptics! *Ironically,* Wikipedia alone is not considered an [authoritative reference](http://meta.skeptics.stackexchange.com/questions/289/wikipedia-as-a-source) for answers. Please add links to more authoritative resources! :-) – Sklivvz Apr 10 '11 at 18:02
  • 22
    @Sklivvz: I would think Wikipedia is an authoritative reference for Wikipedia policies. – David Thornley Apr 10 '11 at 23:57
  • 1
    @David: but ironically it can't be used as such here... – jwenting Apr 11 '11 at 06:37
  • 4
    @Skilvvz I guess you didn't read my post. I did NOT use Wikipedia as reference. I merely used it as example. I hope you'll understand that after you read it. – Bibhas Apr 17 '11 at 06:47
  • That's not what I got from it. I got that the user didn't want to leave two comments. But it's a good point that some might mistake it for that, possibly best to separate the two statements. (I'll do that when welcoming new users myself.) – Goodbye Stack Exchange Jul 10 '11 at 03:07
  • The fact that a wikipedia article cites 'sources' means absolutely nothing unless those sources are credible/reliable. Countless times I have gone to so called 'sources' to find they were simply somebodies blog post, or some referenceless website somebody made in the 90's. – n00b Jan 25 '13 at 23:54
10

I have personally edited incorrect information on Wikipedia. Corporations are known to use Wikipedia as a tool to steer their public image. Sinbad is still among us. Still, wikipedia is great as a preliminary survey on most topics. I use it all the time. Just don't use it as a source on an academic paper and you'll be fine. Your teachers and/or profs are trying to get you to dig a littler deeper, that's all.

Captain Claptrap
  • 828
  • 5
  • 12
  • 1
    Don't use it here either! :-) – Sklivvz Apr 10 '11 at 22:11
  • There are apparently academic sources using it... Look at the line in the question :) – picakhu Apr 11 '11 at 00:28
  • 1
    I've corrected many errors on Wikipedia over time, usually only to see the correction reverted to the previous incorrect information shortly after. There's a LOT of bad data on Wikipedia, and when that data relates to politically sensitive topics, correcting it can create an editing war between those trying to get the truth out and those pushing a political agenda relying on a lie (either deliberately or through ignorance). – jwenting Apr 11 '11 at 06:40
  • 1
    @Sklivvz: Are you serious? I use links to wikipedia often when doing SO. I think most people understand it isn't gospel, but it has reasonable usefulness in making a point. – Captain Claptrap Apr 11 '11 at 21:45
  • @Captain: serious. We have extensively debated the issue on meta, see [here](http://meta.skeptics.stackexchange.com/questions/5/must-all-claims-be-referenced), [here](http://meta.skeptics.stackexchange.com/questions/289/wikipedia-as-a-source), [here](http://meta.skeptics.stackexchange.com/questions/350/downvoting-is-good-for-you-and-for-the-site/372#372), [here](http://meta.skeptics.stackexchange.com/questions/132/list-of-useful-sources-for-skeptics-se), [here](http://meta.skeptics.stackexchange.com/questions/506/what-resources-would-you-use-to-determine-if-somethings-real-or-woo/507#507). – Sklivvz Apr 11 '11 at 21:52
  • @picakhu: No argument here. It is just that the majority of profs and teachers I've come in contact with forbid it as a source in student work. One of their big reasons is that it is superficial and a shortcut to real research. Many also griped about false information, but I personally don't really care too much about that because I'd wager it doesn't average out to much more than what misinfo you'd get in the average non-fiction books at your local library. But who has time to prove me wrong? – Captain Claptrap Apr 11 '11 at 21:53
  • @Captain Claptrap: you hit the nail on the head, the problem is that wikipedia is MORE reliable than other authoritative sources. Why then should professors be allowed say otherwise? – picakhu Apr 11 '11 at 21:56
  • @Sklivvz: So it has been discussed. So what. Obviously you are against it. Ok, but in your 2nd link the accepted answerer made this statement, which I find entirely reasonable: "Linking to such articles should not be discouraged: alternative sources may simply be impossible to come by." – Captain Claptrap Apr 11 '11 at 21:57
  • @picakhu: I'm not sure I agree with that. It is a good goto point for getting started with learning about an unfamiliar subject. If you could rephrase to "wikipedia CAN be more reliable than some other sources" I'll go with you. – Captain Claptrap Apr 11 '11 at 22:02
  • 1
    @Captain: so it's our faq-level consensus here. What *I* want is irrelevant. The community has decided not to accept unreferenced or poorly referenced answers. And - if may add - this question is answerable via *reputable* sources (see the accepted answer). – Sklivvz Apr 11 '11 at 22:02
  • In any case there's nothing wrong with your answer. – Sklivvz Apr 11 '11 at 22:04
8

Wikipedia is an excellent source of general information when starting research. An an example, I'm writing a review of a paper about Leishmania, and the Wikipedia articles that are related and relevant to me (Leishmania, Vaults, 7SL RNA) are quick and easy reads, and loaded with great references to further papers which makes finding them straightforward if not trivial. I might even use the same fact as the Wikipedia article, but it's senseless to cite WP, which in-turn cites the origin of the fact.

Outside of the more scientific topics, however, the amount of references can steeply drop off, which means it won't be of much direct help for any formal research, though it can still help familiarize you with the topics.

Wikipedia articles are not uniformly edited by any stretch of the imagination. Making generalizations about Wikipedia as a whole based on some sort of average quality metric of their articles is dubious.

If anyone (e.g. your teachers) say that "Wikipedia is wrong on point X, as it's Y", that doesn't absolve them of the need to provide some sort of proof or evidence on it as well. Sure, you might be inclined to believe some source above another, but belief shouldn't mix with fact.

Nick T
  • 3,796
  • 3
  • 25
  • 32
3

Here's Wikipedia on The reliability of Wikipedia:

A notable early study in the journal Nature suggested that in 2005, Wikipedia scientific articles came close to the level of accuracy in Encyclopædia Britannica and had a similar rate of "serious errors".[5] This study was disputed by Encyclopædia Britannica.[6]

In areas where I actually have expertise, Wikipedia does quite well. They're not perfect, and you have to watch out for articles that are in the midst of 'editing wars', but if you want to know about RuBisCO or Gamma-ray bursts, Wikipedia does a bangup job.

  • 5
    Bootstrapping answer. Wikipedia is reliable because Wikipedia says it's reliable, and... – Sklivvz Apr 10 '11 at 19:26
  • 4
    It's funny because if you just blatantly copied sections (and their references) from the article, it would have been fine here. – Nick T Apr 10 '11 at 19:34
  • No, it's funny, because it's funny. It also happens to supply a link to relevent information that no one else went to the bother to post. Perhaps the later answer involving the nature study appeared because of this post. That study is after all mentioned in my linked Wikipedia article. –  Apr 10 '11 at 19:51
  • 1
    My original question has that link. – picakhu Apr 10 '11 at 20:18
  • So it does. Sorry I missed it. –  Apr 10 '11 at 23:10