Forgot your password?
typodupeerror
Stats Wikipedia News

Wikipedia-Sponsored Pilot Study Lauds Wikipedia Accuracy 125

Posted by samzenpus
from the accuracy-and-truthiness dept.
netbuzz writes "The Wikimedia Foundation today is releasing the results of a 'pilot study' it commissioned last year to assess the accuracy and quality of Wikipedia in such a way that it would provide a methodology blueprint for others do more thorough reviews of online encyclopedias. The results are in, and despite ready acknowledgment of the small sample size and paragraphs worth of other caveats, the parents of Wikipedia can't help but note that its baby was judged to have outperformed other online encyclopedias, including Encyclopedia Britannica, in three different languages. Britannica, which disputed the Wikipedia-friendly results of a much-cited Fortune comparison report back in 2005, has yet to offer a reply to this one."
This discussion has been archived. No new comments can be posted.

Wikipedia-Sponsored Pilot Study Lauds Wikipedia Accuracy

Comments Filter:
  • by pinkj (521155) on Thursday August 02, 2012 @02:49PM (#40859331)
    In before 'citation needed'!
    • by jhoegl (638955)
      Payment needed before citation needed
      • Re:In before... (Score:4, Insightful)

        by chilvence (1210312) on Thursday August 02, 2012 @04:48PM (#40861011)

        Much as I love wikipedia, a study of the _innaccuracy_ of it would be more enlightening, rather than a study of its accuracy! No one must forget, the fact that the majority of people contribute truthfully to it just makes the minority of people that dont respect it that much more believable... paradoxically, the more faith you have in wikipedia, the more likely that it is to be co-opted... get your head around that! :)

        (that means, I love wikipedia, and I hate lying scumbags, in case you happen to be a lying scumbag and also reading this...)

  • by crazyjj (2598719) * on Thursday August 02, 2012 @02:49PM (#40859333)

    And water wet.

    • This is a known hazard of pulling out a study for your own interests: if it falls against you then you MUST suck, and if it falls in your favor well duh.
      • by sohmc (595388)

        I don't have a problem with companies/organizations funding studies that test their own policies. The problem comes if they in any way influence the results.

        The contract would have to be written basically saying that once the money is received, the benefactor does not have access to the results until published.

        • by KingSkippus (799657) on Thursday August 02, 2012 @03:06PM (#40859553) Homepage Journal

          That doesn't matter. The fact is that if someone approached me and paid a wad of cash for doing something, unless there were some really weird circumstances at work, I'd probably do my best to please them--or at least, to not piss them off--even if they paid me up front and there were "no strings attached." Plus, if you're the company performing this study, you'd have to consider the possibility that the Wikimedia Foundation might want more studies done in the future, if the results you come up with are beneficial to them.

          I've seen this in politics and in corporate studies as well. If at first you don't get a result you agree with, kill the messenger and find someone else to do another study that gives you more favorable results. Bury the first ones and hype the one you like.

          While I'm sure there are some organizations and/or corporations who genuinely want completely impartial results, and there are likewise some companies that generate only completely impartial results, I honestly think that it's the exception, not the rule. Any study should be considered extremely suspect that is directly funded by a company or organization it could benefit.

          • That doesn't matter. The fact is that if someone approached me and paid a wad of cash for doing something, unless there were some really weird circumstances at work, I'd probably do my best to please them--or at least, to not piss them off--even if they paid me up front and there were "no strings attached." Plus, if you're the company performing this study, you'd have to consider the possibility that the Wikimedia Foundation might want more studies done in the future, if the results you come up with are beneficial to them.

            I've seen this in politics and in corporate studies as well. If at first you don't get a result you agree with, kill the messenger and find someone else to do another study that gives you more favorable results. Bury the first ones and hype the one you like.

            While I'm sure there are some organizations and/or corporations who genuinely want completely impartial results, and there are likewise some companies that generate only completely impartial results, I honestly think that it's the exception, not the rule. Any study should be considered extremely suspect that is directly funded by a company or organization it could benefit.

            I think the research groups should commission a study on this. I'd be interested to see the results (or comments on /.)

          • by bws111 (1216812) on Thursday August 02, 2012 @03:21PM (#40859739)

            The 'direct funding' is the key. If you want to have someone do impartial research, hire a third party to hire a researcher for you. That way, you don't know who the researcher is, and the researcher doesn't know who the customer is.

        • by plover (150551) *

          You should be wary anyway. Just picking the researchers introduces a bias. Imagine an oil company funding two studies on seal research, and paying Greenpeace for one study, and Halibuton's House'O'Research for the other. From that point on, they could be scrupulously fair, paying both the same amount, and without asking anything more than "are Arctic seals doing better or worse since the Exxon Valdez incident?" I would suspect the results probably wouldn't correlate well (although it would be interesting t

      • by bws111 (1216812) on Thursday August 02, 2012 @03:18PM (#40859697)

        If companies want an honest opinion about their product (whether it is for PR or for competitive analysis), they should hire a third party. This third party should go and hire the research firm, with the research firm not knowing who the customer is, and the customer not knowing who the researcher is.

    • by Anonymous Coward

      Well, Wikipedia is publishing the methodology and results, so if people have issues with the methodology they can suggest refinements. If they don't have a problem with the methodology, but have a problem with the results, that's *their* problem.

      • by Desler (1608317)

        I think many people will have issues with "the small sample size and paragraphs worth of other caveats".

        • by harrkev (623093)

          Oh, and the "we used Wikipedia as a reference to verify the accuracy of Wikipedia" part.

          >diff wikipedia wikipedia

          Look, it matches!!!

  • by Anonymous Coward

    I bet it has a big old ugly picture of Jimmy Wales across the top.

  • by Chrisq (894406) on Thursday August 02, 2012 @02:56PM (#40859441)
    The quality and accuracy of some articles is great. I would think that most "core" subjects that get a lot of viewers will tend to be of high quality. However look at the entry for so meting obscure, like the town I live in, and you might find something strange. At times there have been mistakes, now corrected - but there is still an odd balance. There is a lot of detail on railway lines that used to go to the station, and what destinations you could reach from the trains.There is a lack of detail on the current geography and economy. Things are driven by people's interests.
    • by CanHasDIY (1672858) on Thursday August 02, 2012 @03:03PM (#40859517) Homepage Journal

      The quality and accuracy of some articles is great. I would think that most "core" subjects that get a lot of viewers will tend to be of high quality. However look at the entry for so meting obscure, like the town I live in, and you might find something strange. At times there have been mistakes, now corrected - but there is still an odd balance. There is a lot of detail on railway lines that used to go to the station, and what destinations you could reach from the trains.There is a lack of detail on the current geography and economy. Things are driven by people's interests.

      Also, any subject (such as, say, Presidential candidates [pcmag.com]) that is/can be politicized is likely to be suspect

    • Yeah, I would say that on articles that have attracted enough attention to have multiple knowledgeable editors, quality is quite good. Exceptions for some rough spots in very hot-button areas, like Israel-Palestine, where sometimes editors with the wrong motives are attracted.

      What I like compared to Britannica is that it's less likely there will be a whopper of an omission in a high-profile article. Some Britannica articles, especially on science/math topics, just have really puzzling stuff missing, or stated incorrectly, while those tend to get found on Wikipedia.

      Of course, they're a bit biased with their list, but a few smug Wikipedians actually maintain a list of Britannica errors that Wikipedia has fixed [wikipedia.org].

    • Precisely - outside of geek culture, pop culture, and the sciences... Wikipedia has some pretty severe quality problems.

      Not to mention the traditional Slashdot lament "who is surprised that a study sponsored by Wikipedia finds Wikipedia is accurate?".

    • by giorgist (1208992)

      That is the key

      Popular topics are better than Britannica because of the many eyes business

      Less popular topics are better than Britannica because Britannica has no article.

  • * According to a study sponsored by the tabacco industry.

  • by ohnocitizen (1951674) on Thursday August 02, 2012 @02:59PM (#40859477)
    Unfortunately the study has been deemed Not Noteworthy by one of Wikipedia's editors and been removed.
    • Thank you for this. That gave me an actually good out loud laugh.
    • by qubezz (520511)

      I'm sorry Wikipedia, but your study must be removed from Slashdot until you can come up with independent citations.

      No original research (NOR) [wikipedia.org] ...."is one of three core content policies that, along with Neutral point of view and Verifiability, determines the type and quality of material acceptable in articles."

  • by Picass0 (147474) on Thursday August 02, 2012 @02:59PM (#40859481) Homepage Journal

    If Wikipedia did this study and kept the results to themselves there would be criticisms regarding transparency. Parties who have a self-interest in damaging Wikipedia would have new ammunition.

  • Very variable in accuracy.

    A few geek-related pages where I have intimate knowledge of the reality, quite often the Wikipedia page is way, way, way off.

    And of course if you make an edit, saying "I WAS THERE, I SPENT 3 YEARS working with that thingy, I HAVE SIX OF THEM IN MY ATTIC", your edit gets removed within an hour, time and again.

    Sigh.

    • Re:Very variable. (Score:5, Informative)

      by Trepidity (597) <delirium-slashdo ... org minus author> on Thursday August 02, 2012 @03:15PM (#40859665)

      And of course if you make an edit, saying "I WAS THERE, I SPENT 3 YEARS working with that thingy, I HAVE SIX OF THEM IN MY ATTIC", your edit gets removed within an hour, time and again.

      That's what they should do, though! Wikipedia isn't a place to publish your own personal knowledge, but a place to publish information that can be cited, ideally to peer-reviewed articles or books. Half the point of a Wikipedia article is being able to look up the references for further reading, and citation where [3] resolves to "[3] Personal experience of Wikipedia user Ancient_Hacker" just isn't very helpful for that.

      • Re: (Score:3, Funny)

        by QuietLagoon (813062)

        Wikipedia isn't a place to publish your own personal knowledge, but a place to publish information that can be cited, ideally to peer-reviewed articles or books.

        If you really believe that, they may not be hope for you.

        Wikipedia is all about certain people taking over articles and, accuracy be damned, making sure those articles reflect the viewpoints of those people.

    • Well you are breaking the rules. This is to ensure accuracy. If you want them to change something, you have to have a reference-able source. So go write the book on it, get a newspaper article published or something, then reference it in Wikipedia as you make the change.
      http://en.wikipedia.org/wiki/Wikipedia:No_original_research#Primary.2C_secondary_and_tertiary_sources [wikipedia.org]

      • Well you are breaking the rules. This is to ensure accuracy.

              I think we know the rules. The issue is that they are foolish. They do no ensure accuracy, in fact, in the case you are responding to, it ensures that is it IS NOT accurate.

              Other than that, well done.

        • Instead of reply, I'll give you a xkcd ... http://xkcd.com/978/ [xkcd.com] :P

        • by Dishevel (1105119)

          Dammit!
          I am right and I know I am right.
          Take your well sourced "truth" and shove it up your ass!
          My word is gold. You just have to trust my IP address that I know for a fact that you are wrong.

          Umm. You know that was sarcastic. Right?

    • by yusing (216625)

      If you have specific objections to the content based on substantial personal knowledge/experience - but don't have time to do the editing (preferable) or to look up a bunch of citations - then bring your points up in the TALK page.

      There they -will- be seen be serious editor(s) and (eventually) will be dealt with. The time this takes can vary from days to years. Faster results may result from snagging an ear in one of the community forums.

  • by jmerlin (1010641) on Thursday August 02, 2012 @03:06PM (#40859555)
    found that cigarettes do not cause cancer. Additionally, smoking cigarettes has the following benefits, probably demonstrated by Tobacco research:

    They increase your coolness factor by 5 points.
    They increase your expected annual salary by 15%
    They increase the likelihood that you will get laid on any given night by 23%
    They decrease the risk of looking like an idiot, since nobody standing around smoking a cigarette can look like an idiot.
    They cause weight gain or weight loss, depending on whether you want to gain or lose.
    They cure the common cold, reduce flu symptoms to 1 day, and potentially cure cancer (we're still checking on that one).
    They inhibit the AIDS virus, no seriously.
    They also increase lung capacity, so if you want to be an Olympic athlete, you should smoke cigarettes!

    In all seriousness, what self-funded studies that find negative things are actually published? You should expect that a headline saying "X funded self study Y" where X is some business that is commonly cited to have some problem and Y is some contention to that commonly held problem. I'm citing this as jMerliN's law of headlines.
    • Incorrect about lung capacity; they do not increase it. Rather, they slowly wither away the lungs, causing the body to adapt to the new changes and therefore conform by retrieving oxygen from other varied sources (e.g. cigarettes). It's the biggest reason why cigarettes are the preferred remedy for those suffering from asthma.

  • This article may contain original research. Please improve it by verifying the claims made and adding references. Statements consisting only of original research may be removed.

  • ...Wikipedia is a hear-say site and this is established by their own policy that limits their liability. In other words: Wikipedia only allows what can be found already published. This was further exposed when through trickery an entry was made that did not yet have other published but upon wikipedia publishing the other made a reference to wikipedia and then the wikipedia article was edited to point to that article as a reference.... And it was found out and the article removed. I don't recall what article

  • I'm going to commission a study to find out why I am so awesome...
    I will keep you posted with the results.

    • by ArsonSmith (13997)

      I don't care why, I just commissioned a study to find out that "I am awesome!" For $50 grand I can do a study, using the same or similar methodology to find out that you are awesome too.

  • http://en.wikipedia.org/wiki/Sega_megadrive [wikipedia.org]

    This redirects to Sega Genesis, even though it was only known as this in North America.

    However, a vote took place on what to call it and as a result of this vote, it's referred to as the Genesis, which is wrong.

    It crap like this that makes me wonder what other facts have been altered.

    • I don't understand what you're saying here. Was the MegaDrive an entirely different product? If it wasn't, then how is the redirection inappropriate? Also, they specifically say in the first sentence, "The Sega Genesis, known as the Mega Drive outside North America". Is that not correct? If you have sources that cite otherwise, why wouldn't you just update the article?
    • How is the information wrong? It has multiple names. The majority of users chose to go with the North American name. The pages still shows the logo for both names, and even makes mention of both names in the first sentence. There has been no altering of facts.
    • by wierd_w (1375923)

      The correct solution to this problem is to regionalize the name, just as the OEM did for the product.

      Eg, for japanese wikipedia, they should refer to the NES as the Famicom, and the SNES as the super famicom.

      The US version should refer the Famicom to the NES, and the SuperFamicom to the SNES.

      Both articles should point out the minor design differences found in the regionalized console offerings. (For instance, the famicom has a microphone built into the controller, and a few other neat things the NES does no

    • Okay, after reading through http://en.wikipedia.org/wiki/Talk:Sega_Genesis [wikipedia.org], it's all fairly clear: you're upset because the consensus on the name didn't match your preference. I don't see why this "makes [you] wonder what other facts have been altered". That's absolutely ridiculous.
      • by metacell (523607)

        The OP is correct. "Sega Megadrive" is the correct name, regardless of what the majority thinks.

        I don't think it's a huge issue, though.

    • http://en.wikipedia.org/wiki/Sega_megadrive [wikipedia.org]

      This redirects to Sega Genesis, even though it was only known as this in North America.

      Goshers, the English edition of Wikipedia refers to the Sega Megadrive/Genesis by the name most native English-speakers called it! That's an outrage!

      • In case you were wondering, the Japanese page is indeed filed under "MegaDrive" (or rather, "ãfããfãf©ããf-").

        • Gotta love Slashdot's total inability to handle UTF. I have no idea how to make actually display the correct katakana.

          • by micheas (231635)

            Gotta love Slashdot's total inability to handle UTF. I have no idea how to make actually display the correct katakana.

            I would have thought the following would work: &#12513;&#12460;&#12489;&#12521;&#12452;&#12502;

            But it displays:

            There must be some weird output filtering going on./p.

      • by metacell (523607)

        The majority of English speakers are not North Americans, if you count all those who use English as a second language. Many of them refer to English Wikipedia since it's the biggest and most well-maintained version, and English is the de facto "international" language in most of the world.

        Remember that English is taught in schools throughout most of Europe, is an official language in India and Malaysia, and so on.

    • by evilviper (135110)

      This redirects to Sega Genesis, even though it was only known as this in North America.

      However, a vote took place on what to call it and as a result of this vote, it's referred to as the Genesis, which is wrong.

      How is the name "Genesis" "wrong" in any sense at all? One sentence up you said it was the correct name used in "North America".

      So the article is filed under "Genesis", there's a working redirect from "Megadrive", and the article correctly mentions the dual-naming issue at the very top.

      There is noth

      • by micheas (231635)

        I agree with everything you said up until " ... and one doomed to ultimately fail under it's own weight."

        The amount of high quality vandalism, for lack of a better term, is apparently, in percentage terms, about the same as other encyclopedias, and Wikipedia seems to catching about half of it. So while it could be a lot better, if you look at the baseline it is an improvement over things like Microsoft Encarta and Encyclopedia Britanica (none of which should be used as an authoritative reference.

        If you ar

        • by evilviper (135110)

          The amount of high quality vandalism, for lack of a better term, is apparently, in percentage terms, about the same as other encyclopedias

          I didn't suggest WP has tons of widespread inaccuracy, I said it will eventually collapse due to vandalism which is being grossly ineffectively dealt with by WP's policies.

          Other encyclopedias don't depend on a dedicated horde of free editors, each offering up a Sisyphean effort to combat all of the downsides of WP's horrible policies. WP has alienated many, many people.

  • Microsoft-sponsored study shows Windows has lowest TCO of all operating systems EVARRRRRRR

  • People aren't mad over Wikipedia's overall accuracy level, which I'm sure is fantastic. They're mad that they can edit any single specific thing to say whatever they want. I don't care if Wikipedia is 99.999999% accurate if I can hit edit and say Barack Obama was born on Mars. It's importance of information combined with ease of editing it that makes Wikipedia lose all respect.
    • This is certainly true, but only half the issue. Wikipedia is justly distrusted because, at any given moment, an article may have been subtly vandalized, astroturfed, tilted in tone, or just plain wrong. Far more important is the ludicrous idea, central to Wikipedia, that any given editor is just as likely to be accurate as any other, without regard to knowledge or experience, that any editor may edit anonymously, and that any system for establishing identity, real-world reputation and (crucially) experti

    • by yusing (216625)

      Try that on a frequently-seen article and see how it takes to get fixed. And (if serious enough) get you banned.

  • All I can say is that I'm still feeling a little guilty about uploading an article about Sgt. Pat McGroyne, who led the charge at the Battle of the Bulge.

A committee is a group that keeps the minutes and loses hours. -- Milton Berle

Working...