Wikipedia talk:Wikipedia Signpost/2015-05-13/Op-ed
Latest comment: 9 years ago by Jayen466 in topic Discuss this story
Discuss this story
- One may better ask, "How can Wikipedia lose its reputation" if the reputation has long been of shoddy, dishonest, frivolous propaganda. Any change in reputation might be seen somewhat more clearly if the light of public opinion polls years apart had been thrown on it. Jim.henderson (talk) 03:03, 16 May 2015 (UTC)
- ... the environment that allowed "anonymous editors" to create the aforementioned incidents has long since dissipated.
- I don't agree. See last month's Washington Post article The story behind Jar’Edo Wens, the longest-running hoax in Wikipedia history for details of a breaching experiment carried out this year: On Monday night, Kohs wrapped up an experiment in which he inserted outlandish errors into 31 articles and tracked whether editors ever found them. After more than two months, half of his hoaxes still had not been found — and those included errors on high-profile pages, like “Mediterranean climate” and “inflammation.” (By his estimate, more than 100,000 people have now seen the claim that volcanic rock produced by the human body causes inflammation pain.)
- The Seigenthaler incident happened right as Wikipedia's popularity was beginning to explode. Wikipedia had about 12,000 active editors in October 2005, a number that has climbed to close to 137,000 now.
- Let's look at the numbers.
- In September 2005, when the Seigenthaler hoax was discovered, the English Wikipedia had a core community of 1,603 editors making more than 100 edits a month (i.e. more than about 3 a day), and 12,201 editors making more than 5 edits a month. Those editors curated a total of 707,000 articles.
- Today, the English Wikipedia has a core community of 3,310 editors making more than 100 edits a month (i.e. more than about 3 a day), and 34,243 editors making more than 5 edits a month. Those editors curate a total of 4.8 million articles.
- So the core community – the sort of editors who might be expected to engage in fact-checking of content they have not contributed themselves – has more than doubled since 2005, but the number of articles has increased almost seven-fold over the same time period, meaning that in real terms, there are far fewer editors per article today than there were in 2005.
- This is borne out by the fact that we still get lots of falsehoods in Wikipedia tripping end users up: Leveson's Wikipedia moment, AFC apologizes to the UAE over ‘Sand Monkeys’ remark on its website, Guilt by Wikipedia: How Joe Streater Became Falsely Attached To The Boston College Point Shaving Scandal.
- Manipulating Wikipedia to Promote a Bogus Business School and The battle to destroy Wikipedia's biggest sockpuppet army, How a raccoon became an aardvark, How I accidentally started a Wikipedia hoax about Amelia Bedelia provide further examples of how anonymous and pseudonymous users are able to corrupt Wikipedia content, introducing distortions and falsehoods that lie undiscovered for years.
- All of the press articles linked here are recent, ranging from October 2012 to April 2015. Andreas JN466 08:35, 16 May 2015 (UTC)
- Good information, Andreas. This trend will continue because there is no limit to how many articles can be created but there is a limit to human resources to manage those articles. Bots etc can solve a lot. The hardest problem is the correct-looking but not actually correct fact intentionally inserted. I think people know this, if not intuitively, which contributes to WP reputation as unreliable. Also, many people who disparage Wikipedia are disgruntled ex-editors who may have been reverted and treated unkindly by those core overworked and surly 3,000 editors - in part our problem is systemically self-inflicted. -- GreenC 15:32, 16 May 2015 (UTC)
- I agree that the superficially plausible fact is the hardest hoax to identify. I do not agree that the number of articles is a problem: the amount of hoaxes is dependant on the number and activity of hoaxers, which at the size Wikipedia has been for some years is more or less independent of additional growth. Therefore the size of the problem is the size of the edit stream. Deeper inspection of edits (or more draconian restrictions on editing) is required to decrease hoaxes. All the best: Rich Farmbrough, 12:40, 21 May 2015 (UTC).
- Surely it's both the number of articles and the number of hoaxers, Rich. Imagine 3 people busy hiding easter eggs in a field: it's your job to find them before a visitor accidentally steps on one. If your field is the size of your living room, with 10 visitors an hour, you can stay on top of things. But if it is the size of a football field, with bushes and hedges blocking your view, and 50 site visitors an hour, you'll find you can't be everywhere. And of course the superficially plausible lie is not always a hoax: sometimes it is just an error, a misunderstanding or an unsuccessful paraphrase. Andreas JN466 01:28, 22 May 2015 (UTC)
- Indeed, for pragmatic purposes the distinction between edits based on motivation is irrelevant.
- But the analogy is largely false: we don't need to be everywhere, just in the recent changes: furthermore the growth in number of articles in the mature project does not correspond to a growth in the amount of bed edits, at least not in a linear way. Someone could perhaps run some stats? All the best: Rich Farmbrough, 12:53, 22 May 2015 (UTC).
- There is no need for a growth in bad edits, and none was stipulated: it's enough for bad edits to survive longer. You're right in theory: staying on top of recent changes would be enough. But that's all academic. As things are, even gross vandalism sometimes gets through recent changes. [1] Moreover, recent changes checking has never approximated anything resembling a rigorous check, incl. verification of sourcing, suitability of added content in article context, etc. It doesn't even approximate that in projects that have pending changes installed (though I believe pending changes cuts down on hoaxes, removing the instant gratification a hoaxer gets from seeing their change go live immediately, and if installed in en:WP would free up time currently spent by RC patrollers on competing with ClueBot). With such holes in the first line of defence, the fact that hundreds of thousands of articles are not on any active contributor's watchlist (or literally not on anyone's watchlist) comes into play. Andreas JN466 18:10, 22 May 2015 (UTC)
- There is no need for a growth in bad edits, and none was stipulated: it's enough for bad edits to survive longer. You're right in theory: staying on top of recent changes would be enough. But that's all academic. As things are, even gross vandalism sometimes gets through recent changes. [1] Moreover, recent changes checking has never approximated anything resembling a rigorous check, incl. verification of sourcing, suitability of added content in article context, etc. It doesn't even approximate that in projects that have pending changes installed (though I believe pending changes cuts down on hoaxes, removing the instant gratification a hoaxer gets from seeing their change go live immediately, and if installed in en:WP would free up time currently spent by RC patrollers on competing with ClueBot). With such holes in the first line of defence, the fact that hundreds of thousands of articles are not on any active contributor's watchlist (or literally not on anyone's watchlist) comes into play. Andreas JN466 18:10, 22 May 2015 (UTC)
- Surely it's both the number of articles and the number of hoaxers, Rich. Imagine 3 people busy hiding easter eggs in a field: it's your job to find them before a visitor accidentally steps on one. If your field is the size of your living room, with 10 visitors an hour, you can stay on top of things. But if it is the size of a football field, with bushes and hedges blocking your view, and 50 site visitors an hour, you'll find you can't be everywhere. And of course the superficially plausible lie is not always a hoax: sometimes it is just an error, a misunderstanding or an unsuccessful paraphrase. Andreas JN466 01:28, 22 May 2015 (UTC)
- I agree that the superficially plausible fact is the hardest hoax to identify. I do not agree that the number of articles is a problem: the amount of hoaxes is dependant on the number and activity of hoaxers, which at the size Wikipedia has been for some years is more or less independent of additional growth. Therefore the size of the problem is the size of the edit stream. Deeper inspection of edits (or more draconian restrictions on editing) is required to decrease hoaxes. All the best: Rich Farmbrough, 12:40, 21 May 2015 (UTC).
- Good information, Andreas. This trend will continue because there is no limit to how many articles can be created but there is a limit to human resources to manage those articles. Bots etc can solve a lot. The hardest problem is the correct-looking but not actually correct fact intentionally inserted. I think people know this, if not intuitively, which contributes to WP reputation as unreliable. Also, many people who disparage Wikipedia are disgruntled ex-editors who may have been reverted and treated unkindly by those core overworked and surly 3,000 editors - in part our problem is systemically self-inflicted. -- GreenC 15:32, 16 May 2015 (UTC)
- Per "Wikipedia is an encyclopedia written by, for the most part, laypeople". For medical content I would disagree. The majority of our medical content is written by health care professionals. [2] And many Wikipedia editors are experts. Doc James (talk · contribs · email) 10:01, 16 May 2015 (UTC)
- About the headline: somehow I think it might be better if we played to the viewer's opinions of themselves, and change it to "the free encyclopedia that you can edit", like French Wikipedia currently does. (I first found this idea on Basemetal's userpage, but then I noticed that some other WPs actually do use that amended headline.) Double sharp (talk) 10:46, 16 May 2015 (UTC)
- Wikipedia will never be what it wants to be, a solid source of reliable info, because it lets the masses edit and will always be incapable of putting sufficient control mechanisms in place to control the bullies, POV pushers, and outright mentally unbalanced users who have far too much control here. HalfGig talk 10:55, 16 May 2015 (UTC)
- It's unclear to me what the source is for the statement that there are currently 137,000 active editors. Per the official statistics for Wikipedia, there just under 76,000 active editors in March 2015, across all language versions. That's down about more than 15% from the peak in 2007; it is also sharply up from the 2005 figures. (To be specific, March 2007 had just under 89,000 active editors, compared to just under 51,000 in March 2006 and about 14,000 in March 2005.) So yes, the number of active editors has sharply increased since 2005, but essentially all of that increase occurred between 2005 and 2007; the count has been generally downhill since then. -- John Broughton (♫♫) 04:29, 17 May 2015 (UTC)
- I think that one fact that is easily overlooked is that the number of former Wikpedia editors (say, who haven't edited in three years or more) outnumbers the current number of active editors. And it might outnumber the very active editors by a hundredfold! So, there are a lot of people who speak of Wikipedia with familiarity who were either blocked or got disillusioned or chased off or just got busy with life and quit editing. In a forum like Quora, a person who answers a question about Wikipedia could have, say, edited intermittently between 2005 and 2007, had a few bad encounters and left. But given their participation, I imagine they feel as qualified to answer questions about Wikipedia on Quora as an enthusiastic new editor who has been here for 4 months and thrown themselves into the project.
- Meaning, because Wikipedia is such a highly ranked and familiar website, there is an incredibly varying level of experience and knowledge about it among individuals, and an appreciation of how it has changed over time really probably is only apparent to a small sliver of editors and readers. Liz Read! Talk! 20:17, 17 May 2015 (UTC)
- I think the 'Don't use Wikipedia' statement is misunderstood: I don't want students to use it because it is tertiary information, an encyclopaedia. They should use papers and monographs instead. Usage of Encyclopaedia Britannica is likewise improper in an academic context. That has nothing to do with trust. --Pgallert (talk) 08:55, 18 May 2015 (UTC)
- Thank you for your comments. Just yesterday one of my brothers complained about the misinformation that keeps popping up in the English Wikipedia article on Adam DeVine as evidence that you can't trust Wikivoyage. I know -- it's a logical error -- but that's what people think. One reason I got involved in Wikipedia was the over-use by my students. I solved that problem by limiting students in their citation of tertiary sources: to one (in a 200-level course) or two (in a 100-level class). Bearian (talk) 11:54, 18 May 2015 (UTC)
- Standards for sources seem pretty low to me. There's a reason legitimate news organizations don't cite many of the publications wikipedia will happily accept as a source. — Preceding unsigned comment added by 94.66.218.131 (talk) 13:52, 19 May 2015 (UTC)
- More simply, Wikipedia suffers any number of people (celebrities, media, etc.) upset about inaccurate articles without either learning why the inaccuracy exists or doing anything about it. Wikipedia doesn't have a chorus exalting how good our articles are. We only have readers that show up, get the content they want, and leave. Many readers are impressed with Wikipedia until they find clickbait that tells them otherwise. I don't think beating the drum about how n00bs get treated is germane here. Chris Troutman (talk) 00:40, 20 May 2015 (UTC)
← Back to Op-ed