Wikipedia talk:Improving referencing efforts

From Wikipedia, the free encyclopedia
Jump to: navigation, search
2007 discussion archive

Barnstar incentive[edit]

Please discuss the barnstar incentive proposal here.

Educate editors[edit]

Please discuss how to educate editors here.
Should we have templates to inform editors they added articles without references?
Personally, I think having templates to inform editors that they have added an article without references is fine, and would not hurt in an effort to improve referencing efforts. However, I think their impact would be low (i.e. IMHO I think people would simply ignore them). I know this is anecdotal, but from my own experiences working with editors at WP:DERM, repeatedly commenting on an editors talk page regarding lack of citations has had little effect, so I feel a template would get about the same, or worse, response. kilbad (talk) 12:14, 12 December 2008 (UTC)

If you are looking for useful essays, you can always point people to User:Uncle G/On sources and content#Tips for editors and User:Uncle G/Wikipedia triage#What to do. Uncle G (talk) 16:27, 6 December 2008 (UTC)

The first page mentioned was viewed 22 times in November 2008. The second was viewed more - 89 times in November 2008. By comparison, 150,000 registered editors made one or more edits in the past 30 days. I'm not against educating editors (sure, why not?), but we need a more effective approach than to modify these two rarely-viewed pages. -- John Broughton (♫♫) 21:06, 28 December 2008 (UTC)
  • Better to educate interested editors on how to find and add citations faster. (talk) 03:52, 5 January 2009 (UTC)

Adding a citation template button to the toolbar[edit]

Please discuss the citation template button proposal here.
I use a firefox add on tool called WP Cite, which I think is more useful than an add on to the Wikipedia website, because it avoids having to copy and paste each item of information. The tool needs to be updated for Firefox 3 though. PhilKnight (talk) 00:26, 8 December 2008 (UTC)
  • Problem is, not everyone uses or wants to use Firefox. (I don't have admin access on this machine, so I'm stuck with IE) Also, I occasionally run across things that make Firefox break. I want to be able to use alternatives. - Mgm|(talk) 09:52, 9 December 2008 (UTC)
In the Gadgets section of Special:Preferences, logged-in users have the ability to easily add User:Mr.Z-man's Cite button to the editing toolbar. This useful tool gives users the cite template parameters and automatically formats them. This may be the same tool, only a different way to implement it. This button should be enabled to all users, not just those logged-in who checked the box. Reywas92Talk 22:50, 16 December 2008 (UTC)
  • Support if it functions I like it. Citations are a pain to add correctly, and I have nearly given up on Wikipedia because of it. Any tech fix will be welcome. (talk) 03:52, 5 January 2009 (UTC)

Press release[edit]

Please discuss the press release/cleanup proposal here.
There are certainly dozens of different improvement efforts underway (dozens of cross-cutting WikiProjects), all of which, I'm sure, would like to issue a press release in order to attract more editors. But at Wikipedia, I'm fairly sure that things have to be done in the reverse order: attract (somehow) a large number of editors who are willing to participate, and then do some (more) publicizing. We also should keep in mind that only the Wikimedia Foundation has the authority to issue official press releases, and they are very unlikely to do so for a single language version (English) of a single project (Wikipedia). We'd probably have to settle for an article in the Signpost, and a bit of posting on en.wikipedia-related mailing lists. -- John Broughton (♫♫) 21:12, 28 December 2008 (UTC)

Bots and scripts tagging new pages[edit]

Please discuss the bots and other technical proposals here.
Should/can we use bots to sort unreferenced entries from the incoming articles?
Should/can we use bots to sort unreferenced articles by topic and leave them at relevant userpages and wikiproject pages?

I think a bot that posted a list of unreferenced articles to subpages of relevant WikiProjects could go a long way toward getting more editors involved in fixing this problem. (The bot should also drop a note on the main WikiProject talk/discussion page whenever such a subpage is created or updated.) With such a bot in place and running, the issue would then be that unreferenced articles often don't have any WikiProject templates on their article/talk pages. Without such templates, a bot can't do anything. If we had such a bot, then we could focus on how to build a group of editors to do talk page tagging (which is valuable in itself), perhaps using some sort of script in order to divide the work (for articles that have one or more categories). -- John Broughton (♫♫) 01:25, 28 December 2008 (UTC)

Stable versions[edit]

Please discuss stable versions here.
Can someone point to some details regarding this assertion? My understanding is that stable versions will not prevent any editing; it will only help ensure that edits are reviewed. But if the criteria for speedy deletion of articles remains unchanged, and editors are no more willing to take unsourced articles to AfD than they are right now, and reviewing editors don't feel any responsibility for tracking down sources for unsourced articles (should they?), then it's not clear that stable versions (flagged revisions) will actually make that much difference. (They can make a difference in which versions are presented to readers, but that's another matter, more related to vandalism.) -- John Broughton (♫♫) 21:16, 28 December 2008 (UTC)

If an article needs references ... add them[edit]

Please discuss If an article needs references ... add them here.
    • First, the main problem being discussed here seems to be unreferenced articles, not unreferenced statements. Second, if a statement is false or is true but comes from personal knowledge, it will be impossible to find a source; one is looking instead for a source that contradicts the statement, which is much harder. Third, a large number of Wikipedia articles are on technical subjects; citations for (for example) a mathematical theorem are likely to be beyond the capabilities of a large percentage of editors. Fourth, while it's true that an experienced editor can often find a source and create a citation in less than a minute, it's also true that experienced editors often find that they can't easily find a source, and that inexperienced editors can find the citation and footnoting requirements to be beyond what they are willing or able to learn.
    • And even if it's true that the problem is simply an attitude of unwillingness to accept responsibility, do you have any suggestions for changing this attitude? (Because if you don't, your pointing this out is in no way a solution to the problem that everyone agrees exists.) -- John Broughton (♫♫) 21:47, 1 January 2009 (UTC)
      • From my experience, your estimate is plainly wrong. For somebody who adds information, it may in some cases take 30-45 seconds to add source information. (If he knows what source he is using, that is.) But for someone else who is trying to verify existing information, 30-45 minutes is often a more accurate estimate. (To find a source that actually supports the content, not to provide a link to Google search results, that is.) Quite apparently, nobody is willing (or even able) to invest this effort. The burden of evidence is with the editor that adds information, and that's for a reason. Allowing to add unsourced information, and leaving it for others to clean up, does not lead us any further, as the current situation clearly shows. --B. Wolterding (talk) 13:20, 1 January 2009 (UTC)

Add links to sources[edit]

Please discuss Add links to sources here.
    • A bot might do this automatically. Given that 6,000 to 8,000 articles per month are being tagged with the {{unreferenced}} template, it would seem to make sense to have a bot automatically add the {{find}} template to the related article talk pages. If the bot also notified the editor who created the article that (a) it had been tagged as lacking sources, and (b) the talk/discussion page had a helpful link for finding sources, then there would be a greater likelihood that the article's author would help fix the situation. -- John Broughton (♫♫) 00:48, 28 December 2008 (UTC)
PhilKnight (talk) 13:28, 30 December 2008 (UTC)
As for the bot adding the {{find}} template, I'd guess that most of the unreferenced articles have fairly unusual titles - that is, aren't things like "John Smith". It would help, certainly, if the bot were smart enough to take an article on Fred Jones (footballer) and create the find template with two parameters ("Fred Jones" and "footballer"). -- John Broughton (♫♫) 01:55, 31 December 2008 (UTC)
  • Support That {{find}} tool rocks. It should be included pretty much everywhere. (talk) 03:52, 5 January 2009 (UTC)
  • Comment. Perhaps instead of just dropping {{find}}; we have a modified template that explains how to adjust the search strings for it. On most articles I use several search strings to find various aspects of a subject. -- Banjeboi 11:55, 10 April 2009 (UTC)

Research Honours - Research Apprentice / Research Master / Research Doctor[edit]

Please discuss Research Honours - Research Apprentice / Research Master / Research Doctor here:
  • Better to train users who are good editors but don't do much citing to do a better job. (talk) 03:52, 5 January 2009 (UTC)

Simplify the process of adding citations[edit]

Please discuss Simplify the process of adding citations here:
anyone can add citations in any format at all. Somebody will fix them. It's just like grammar & spelling and the like. It's better to spell correctly, and use correct grammar, but there are many people who like to take care of problems here. Myself, I tend to use the simplest clear format and go on to find another. Anyone who chooses to elaborate it into a full {{cite}} is welcome. DGG (talk) 00:34, 6 January 2009 (UTC)

New CSD: Newly created unreferenced articles[edit]

This is the 2008 village pump thread that caused me to reopen this discussion.

The categories with unreferenced articles and articles with unreferenced statements are growing faster than editors can clean them up. In order to catch up, I would like to propose a speedy deletion criterion that allows speedy deletion of articles created after the CSD was accepted and was left unsourced for at least x days. The number x can be debated. It would encourage people to not leave their articles unsourced and it would allow editors to catch up with referencing material.

A similar strategy was applied to non-free images uploaded after a date some 2-3 years back to stop the flow of unfree images.

As I imagine it, it would only apply to newly created material that is fully unsourced. Material with dubious sources or flat out bad ones should go through AFD.

Is there any sort of support for an idea like this? - Mgm|(talk) 11:36, 5 December 2008 (UTC)

This and similar proposals have arisen previously and been rejected, though I support such a measure (and more extreme) wholeheartedly. See Wikipedia:Requests for verification. The hole we have dug for ourselves by not requiring sourcing under any pragmatic standard is very, very deep. There are many hundreds of thousands of unsourced articles and climbing every day. It is a cancer on Wikipedia and m:eventualism is a pipe dream.--Fuhghettaboutit (talk) 12:52, 5 December 2008 (UTC)
  • I don't think eventualism itself is a problem. We can't expect articles to be FA quality from the beginning. Applying eventualism to references is what is particularly problematic. - Mgm|(talk) 13:01, 5 December 2008 (UTC)
  • There seemed to have been a general idea that instead of deleting, you should provide sources. Others disagreed and said the burden was on the creator. If this is to succeed, those two need to be reconciled. - Mgm|(talk) 13:13, 5 December 2008 (UTC)
  • Apparently a lot of editors don't notice at the bottom of the edit box, and I can't fault them. It's an obscure link and no one warns them when their content is not verifiable. Perhaps we should first focus on warning people who create new articles without any references about the problem they're creating. - Mgm|(talk) 13:27, 5 December 2008 (UTC)
  • Or both. I support CSD. dougweller (talk) 13:51, 5 December 2008 (UTC)
  • This proposal really catches my attention, but it makes me uncomfortable. On the one hand, requiring sources for all new articles is difficult—it raises the bar higher for new contributors, and feels like bad faith to me: are articles really "non-notable until proven notable"? On the other hand, this is already largely the case: unsourced articles get deleted, especially anything that already falls under CSD A7. Codifying it might make it simpler to explain what we as a community expect, which could be friendlier to new users. Do we raise the bar and make our rules clearer, or do we leave it where it is and allow new users some leeway? On a similar note, would this CSD apply retroactively to existing articles? Applying this retroactively would not be a good idea. {{Nihiltres|talk|log}} 14:16, 5 December 2008 (UTC)
  • Er, no. If the article is weak and unreferenced send it to AFD or PROD it. I see very little difference between this and those two already approved existing options.---Balloonman PoppaBalloon 16:47, 5 December 2008 (UTC)
  • No, WP:V does not require an article be referenced, it requires an article be verifiable. The 2 are different things. Just because people would rather remove content than spend 30 seconds googling for a ref does not mean its unverifiable. Technically only material likely to be challenged needs a ref. I guess we could just say "all new material is automatically suspect" - but that's kind of dickish. I'm about as deletionist as one can get, and I still think this is too far, deleting new unsourced BLPs might be reasonable, but its unnecessary for all articles. If we do this, we should forbid new users from creating articles too, because otherwise they're all just going to be deleted. Mr.Z-man 17:14, 5 December 2008 (UTC)
  • Oppose. Use WP:PROD if you think the unreferenced article needs to be deleted (in many cases they need to be improved rather than deleted and anyone can do that, even you  ;-) ). I don't think it makes sense to extend speedy deletion to things like this. Dragons flight (talk) 17:21, 5 December 2008 (UTC)
    Agreed---We already have enough overly eager CSD'ers out there mistagging articles wrong and antagonizing people. CSD is a necessary evil, but CSD'ers can do more harm to the project than the worst vandal. I do not like the idea of increasing the amount of unmonitored damage they can perform. AfD exists for a reason, use it.---Balloonman PoppaBalloon 17:35, 5 December 2008 (UTC)
  • No, most of unreferenced articles are harmfulless. They can be improved and referenced, while deleting them would extraordinarily damage our coverage. When they really need deletion, it's possible to prod them in non-controversial cases, afd them in other cases. And we have G10 for unsourced negative blps. Don't forget that most articles on places, to a lesser extent on organisms and other typical categories, are unreferenced but constitute a large percentage of our articles. Non-free images are dangerous for the project, but not those articles. Cenarium (Talk) 18:56, 5 December 2008 (UTC)
    I think you meant: No, most of unreferenced articles are harmfulless---Balloonman PoppaBalloon 19:19, 5 December 2008 (UTC)
    Yes, thanks for pointing that out. Cenarium (Talk) 19:25, 5 December 2008 (UTC)
  • I used to think this would be a good idea, but since then I've done a bit of work referencing articles in these categories and changed my mind. There are quite a few articles in these categories which do end up getting deleted, but the majority of them are perfectly good articles apart from the referencing issue. Take this example - this person served as attorney general and chief justice in several countries and so is clearly suitable for inclusion, but it doesn't cite any references. A badly referenced article is better to the reader than no article at all, and it is a better place to start from when building a better one. Furthermore few new articles cite references, and those that do are generally written by more experienced editors, so this would be a serious bar to newcomers. Hut 8.5 16:47, 6 December 2008 (UTC)
  • Not sure this is necessary - I don't think anything more than {{prod-nn}} is required. PhilKnight (talk) 00:29, 8 December 2008 (UTC)

Alternative proposal[edit]

  • What could be done, instead of deleting unreferenced new articles, would be to userfy the article to the userspace of the article creator, with a note made to their talk page advising them of the need to add at least one reference before it can be moved back to mainspace. A list (or category) could be made of such userfied articles and if they are not referenced and moved back to mainspace within a period (1 to 2 weeks?) it would then be speedily deleted. This is less bitey to newcomers, preserves the content that they may have spent a while creating, while not leaving them indefinitely in either mainspace or userspace (there is a certain time for the user - or anyone else - to add a reference). This would, of course, not prevent existing speedy deletion candidates from taking place. Davewild (talk) 16:41, 5 December 2008 (UTC)
I agree - userfy, warn then de-userfy when referenced. I don't agree with deleting them though - people, especially newbies, won't want to focus on finding sources for one article. In userspace it's fine. Dendodge TalkContribs 16:46, 5 December 2008 (UTC)
I've read many useful WP articles that have been totally unreferenced - sometimes (very often?) people just come along and write about something they know about. In fact, does it really make that much difference if someone adds one purported reference (which might be irrelevant, unreliable or even made up)? Better to allow good sources to be found over time than to coerce people into adding bad ones at the outset.--Kotniski (talk) 17:06, 5 December 2008 (UTC)
I agree with you 100%. I have been having this problem with a project I started, WP:DERM. I have really been pushing new contributors to cite sources, but by pushing/coercing them, they simply add bad ones as a "quick fix." It has really been frustrating, so I am a believer now that is it better to allow good sources to be added over time. kilbad (talk) 01:26, 10 December 2008 (UTC)
again, disagree. If the article needs to be deleted due to lack of references, send it to AfD. it does the same thing, but brings it to other people's attention who can perhaps add resources. It might also be a notable article, but simply lacks resources. CSD doesn't address this.---Balloonman PoppaBalloon 17:33, 5 December 2008 (UTC)
Disagree also. Articles are supposed to be edited, improved, referenced, etc, by any editor, this is the strength of Wikipedia. If they are put in userspace, they'll be forgotten most of the time, while they could have been improved while in mainspace. Unreferenced articles are generally not harmful, and {{unreferenced}} is there to alert the reader. Cenarium (Talk) 18:32, 5 December 2008 (UTC)
Those proposals would be incredibly harmful to the scope of our project. Again, we have tens of thousands or unsourced articles that are still highly beneficial for Wikipedia, and they always can be checked for verifiability and then acted upon. That they are old or newly created doesn't change that. It would also discourage new editors that we lack so badly. Cenarium (Talk) 19:09, 5 December 2008 (UTC)
We do not lack in new editors at all. We get plenty of those. - Mgm|(talk) 11:09, 6 December 2008 (UTC)
I have seen many users create an article, then never come back. If their article was userfied, it would pretty much be lost forever. At least if it's in mainspace, other people may come across it and improve the article in time. Somno (talk) 04:55, 6 December 2008 (UTC)
Yeah, this is particularly true if say userA writes a B-class article, it just lacks references. It is userfied, but userA never returns. UserB writes the same article, but it is only a start class article---but it has references. The project would have benefitted if UserB could have added references to UserA's articel, and perhaps moved it to GA quality.---Balloonman PoppaBalloon 07:14, 6 December 2008 (UTC)
  • What strikes me as odd is that we have a CSD criterion for such a fuzzy thing as notability (it's such a fluid concept that people rarely agree. If we oppose forcing people to provide sources, which I think is a very bad idea, at least we should make them aware of the work they're causing others and ask them to fix it. - Mgm|(talk) 11:09, 6 December 2008 (UTC)
  • Yes, technically we can reference unreferenced articles, but there are several issues:
  1. It requires expert attention of the subject.
  2. It's time consuming
  3. While we can, statistics show that the people who actually do it, are outpaced by the people who add unreferenced material.
  4. Leaving unreferenced material lying around means it gets copied across the net to the point we don't know what was there before the article was written.

This leads me to three conclusions. a) Either we need to increase the number of people adding references to unreferenced articles somehow (which despite efforts doesn't seem to be popular) b) Stop the flow of incoming unreferenced material. Since deleting it doesn't has a consensus, the obvious next idea would be to get editor themselves to sort it. A lot of material is taken from a source rather than memory and at least having that info available (even if the source is unreliable, makes things a lot clearer) c) An article should be referenced sooner rather than later, to avoid it being referenced with copies of itself.— Preceding unsigned comment added by MacGyverMagic (talkcontribs)

Wouldn't tagging the article as unreferenced and putting a note on the author's talk page explaining verifiability be sufficient for "getting the editor themselves to sort it"? Of course, they might not do anything about it, but that's true for everyone and everything on Wikipedia. Also, referencing does not require expert attention - anyone can do a Google search or go to the library to look something up. It only requires willingness to do that, not necessarily expert knowledge. Somno (talk) 00:58, 8 December 2008 (UTC)

Encouraging WP:BOLD and widening RS criteria[edit]

I think WP:BOLD goes hand-in-hand with some of the proposals being put forth here. One thing that should be considered is the onus should actually be removed from the creator/contributors to an article and instead placed on all editors to add sources. What I mean by that is there are far too many cases of "drive-by tagging" and "drive by AFD'ing" going on -- editors jump onto an article, see no references or no references they like (more on this in a moment), and so they tag the article or peg it for PROD or AFD, and move on, having made little effort to actually add sources themselves. Of course many editors DO indeed do this, and many AFD's are nominated by editors who have at least made an effort to find sources. So I'm not intending to generalize. But I think the message should go out that if you find an article lacking sources, just as editors are encouraged to fix typos and other errors and remove vandalism on sight, so too should they be encouraged to at least make an effort to find sources, if at all possible.

On a related note, Wikipedia needs to update its policy regarding what sources are acceptable. The bias against blogs, fansites, and other Web-based content needs to stop, especially when such sources are fast becoming recognized as sources of record, even without the existence of professional oversight. This was acknowledged by no less than CNN during the Mumbai attacks, when it recognized that the blogs and Twitter and other online sources were ahead of tradtiional media in reporting the event. This is, of course, a matter for another discussion, but I think sourcing of articles will improve once Wikipedia broadens its definition of WP:RS. At the same time, however, I also see the occasional bias against non-Web sources. I've seen the occasional AFD where the nomination has been based upon lack of Internet sources, only to have defenders of the article cite printed sources that have never been put online. The bias can go both ways, and this too needs to be addressed if we want better referencing/sourcing on our articles. 23skidoo (talk) 16:20, 10 December 2008 (UTC)

  • The onus has never been on the creator/contributors alone in the first place. See User:Uncle G/Wikipedia triage#Looking for sources yourself beforehand. The onus has always been on editors to also look for sources themselves when challenging verifiability. This has been verifiability policy since its inception.

    When people write things such as this, they are, quite simply, completely wrong. That is not, and never has been, expected behaviour of editors here. It should be obvious that such an attitude is unconstructive, unhelpful, uncollaborative, and not conducive to the actual betterment of the encyclopaedia. Any editor asserting this does not understand how our verifiability and editing policies interact, and requires education in how we have always collaboratively worked towards keeping content verifiable, helping one another with sources in (yes) the same way that we help one another with typing errors.

    Writing the encyclopaedia is not Somebody Else's Problem.

    And the problem with WWW content is not ours to fix, and certainly not a reason to abrogate our standards. The problem stems from the fact that much such content is not associated with identifiable people who have reputations for fact checking and accuracy. Pseudonyms and lack of identification are why readers do not trust us, and are similarly why readers cannot trust many other WWW pages, and pretty much all Usenet postings, as well. Much of the WWW is, quite simply, not published in such a manner that any reasonable person can trust its accuracy and veracity, or by people who have any expertise in what they write. Ironically, our motto of "making the Internet not suck" derives in part from the fact that a lot of it does, in this respect. Uncle G (talk) 03:23, 14 December 2008 (UTC)

  • There are too many people with a bias against blogs. Blogs aren't neccesarily unreliable. As long as the blog is attached to an identifiable individual who is supposed to be an expert. Fansites tend to spread rumors and forum posts are hardly ever positively linked to a person. So I wouldn't accept those in a hurry, though there is a specific forum I'd like to use as a reference. Problem is, the reliable person involved can't confirm they're passing on the information :( - Mgm|(talk) 11:58, 15 December 2008 (UTC)

Old unreferenced articles[edit]

A large part of the problem is that before 2006 sourcing wasn't widely enforced; the {{fact}} tag didn't even exist until the summer of 2005. Thus many articles created before then don't have references, especially if there on somewhat obscure topics. When old articles show up at AFD, often this is the reason. (talk) 16:28, 12 December 2008 (UTC)

Are we sure this is a problem?[edit]

Periodically, I have done a little test. I hit random ten times, and see how many of the articles have any referencing. A year ago, I usually would find about 0 out of 10 articles have references. I did it just now, and got 4 out of 10 articles with some sort of referencing. Some of those articles had more than one ref, so the average was about 1 ref per article. I think over the past year we have actually been adding refs at a very fast pace, probably 1 ref/article/year. Has anyone ever done an analysis. Maybe count the average number of <ref.> tags from a year ago and today. I think you will find that they are increasing quickly. - Peregrine Fisher (talk) (contribs) 01:03, 15 December 2008 (UTC)

  • If six out of ten articles have no references whatsoever, there is a serious problem, yes. If the vast majority of the content of an encyclopaedia is not supported by secondary sources, it does not speak very well for the integrity of the encyclopaedia. Skomorokh 01:09, 15 December 2008 (UTC)
    • It's not the number of pages with or without refs, it's how fast are they being refed. If each year, 4 more out 10 are refed, then in 1.5 years all articles will have a reference. That's probably not how the trend really is. My guess is that 1 year ago our articles contained an average of about 0.1 references. Six months ago it was about 0.3. Now it's about 1.0. Who knows if this is a linear of exponential curve. If it's the first, the average article will gain about 1 ref a year. If its exponential, we may end up with an average of 10 refs per article in a few years. The emphasis on refs hasn't been around that long, so we should quantify if people are solving the problem or not. - Peregrine Fisher (talk) (contribs) 01:23, 15 December 2008 (UTC)

For any statistics here to hold water, and to support meaningful conclusions, one needs a larger sample size than a mere ten pages out of a population size of more than 2 millions. Uncle G (talk) 01:49, 15 December 2008 (UTC)

I don't disagree that my sampling isn't the best. I just did another test. I got 18/32 articles selected using random (if random isn't really random, then this is meaningless). Using this confidence interval calculator, with a 95% confidence level, between 39 and 73% of articles currently have references. More than I thought. I didn't inlcude dab pages. Do you know of any way I can try this on a past version of wikipedia? - Peregrine Fisher (talk) (contribs) 02:05, 15 December 2008 (UTC)
Use the history function. Juzhong (talk) 17:18, 15 December 2008 (UTC)
I'm trying to find an easier, more accurate way. Good suggestion, though. I've been asking for help on this on various pages. I got this response over at Wikipedia:Bot requests/Archive 24#Compile statistics on article referencing.
I did something similar a few months ago, using all articles in the static html dump from June 2008. By counting the number of <ref></ref> tags I found that the average number of citations per paragraph was 2.07 for FA, 2.06 for GA, 0.87 for A class, 0.51 for B, 0.26 for Start, 0.14 for Stub and 0.15 for Unassessed articles. This was out of a total of 2,251,862 articles (disambig pages and obvious lists excluded); 1,625,072 (72%) had no <ref></ref>s, though some of these will have Harvard-style refs, e.g. (Smith 2000). Dr pda (talk) 23:46, 16 December 2008 (UTC)
It doesn't show our trajectory, but it's pretty interesting. All we need is a good analysis of our current state, and we can start seeing how good/bad of a job we're doing. - Peregrine Fisher (talk) (contribs) 02:42, 17 December 2008 (UTC)

Wikimedia received a grant to create an easier editing experience[edit]

I imagine that making referencing easier would be on of the things they would focus on. I wrote an article for the Signpost about it at User:Cryptic C62/Workspace. It's the part titled "Wikipedia receives grant to make editing more user-friendly". - Peregrine Fisher (talk) (contribs) 02:15, 15 December 2008 (UTC)

Conflict between reference and original research requirements[edit]

As a well-meaning contributor with relatively limited resources, I have become increasingly frustrated by the (from my anecdotal perspective at least) overwhelming tendency of editors to "raid and destroy" information rather than attempting to improve the existing (albeit unsourced) material. My experiences in this regard have been sufficiently disheartening to make me actively consider ceasing to contribute. Thus I have given the "dismissive deletion" issue much thought in the last few weeks (feel free to check out my various village pump contributions if you're a true glutton for punishment−I would provide links, but they get archived so fast the links would probably be broken anyway), and I would like to discuss one of the patterns I have noticed in the course of my ruminations.

The original research requirement is one of Wikipedia's oldest policies, and was apparently instituted to make it harder for contributors to abuse their edits to advance a non-neutral point of view, especially in a scientific context. As the official history has it, "the original motivation for NOR was to prevent editors from introducing fringe views in science (emphasis mine) ... or from excluding verifiable views". This is all very well for scientific topics, where researching sources is comparatively easy. For a great many nonscientific topics, however, the original research policy is so problematic as to become nearly paralyzing to a well-meaning editor like myself. On the one hand, contributions that have taken considerable effort on the part of one or (in many cases) a considerable number of editors are routinely destroyed by hit-and-run deletionist editors for lacking source citations. On the other hand, for many, many nonscientific topics, especially in entertainment and popular culture, where the very best sources available are often "blogs, fansites, and other Web-based content" (as another contributor put it above), the provision of a "reliable" source is not only difficult, but becomes a dishearteningly easy target for deletionist editors to kill based on the original research requirement.

Consider [[1]] this edit from the Casino Royale (1967 film) article. The deletion in question is typically dismissive, with no more than a perfunctory note in the edit summary. Significantly, the dismissive deletion destroys a link to the relevant director, effectively without replacement. The damning accusation of original research is (again typically) taken for granted. As it happens, the information can be checked by watching the film The Life and Death of Peter Sellers just mentioned higher in the same paragraph, which is, of course, a "primary source". However, by this reasoning it is not at all clear that it would not constitute original research! By the same logic, every single reference whatsoever can potentially be ruled out of court. After all, if it's online it's suspect and unreliable, and if it's from a book or journal, someone would have to go to a library to look it up!!! Thus, the typical draconic dismissive interpretation of the original research rule effectively discourages editors from looking for sources. It takes time and effort to look up these references (which are obviously required to protect the information from being unceremoniously destroyed). However, providing a source may not only not be enough, it may actively endanger the information all over again. For the record, the information deleted in the above example can be found in Chapter 12 on the 2005 HBO DVD 92833A, time stamp 1:24:15 to 1:25:24. Who wants to bet the information would be safe if this source citation were added? And yet the information does not try to advance any opinion or point of view, it can be verified by anyone willing to watch the DVD.

I could give other examples (and I have done so several times elsewhere), but I have been told before that people ignore me for being too long winded, so I'll just summarize for now: I fear that in a great many nonscientific fields, editors may be discouraged from providing references because the original research requirement tends to be interpreted too broadly. Nude Amazon (talk) 06:13, 15 December 2008 (UTC)

References != accuracy[edit]

If you look at my edits, most anything I add to articles has a reference. But the little superscript blue number at the end of a sentence means remarkably little... as much I'm for references, I'm for verified references. People still need to actually check these things for them to mean much. And from verifying real world academic article sources, Wikipedia would hardly be the only place where sources are added for decoration. I've had college professors, when pressed, all but admit they made up some citations to make a paper look better, in some cases the article they referenced didn't even exist, let alone back up the claims it was supposed to. It's rather disturbing.

References are nice, but unfortunately, a good article needs more work than just a bunch of ref tags. It takes someone actually spending vast amounts of time making sure the claims correspond to actual sources. There's no really easy way to make that happen. We like to think we can just add a tag or spend 60 seconds and make an article nice and verified... but unfortunately getting an article right is an ordeal that really can't be simplified. There's only one way to do it and that's the hard way.

Anyway, what's my point? Just kind of thinking aloud for the most part. Adding references is still a useful step, since it gets us started on the road to verifying claims. But unfortunately adding references is just the first step, not the end of the process. --Rividian (talk) 04:18, 17 December 2008 (UTC)

The current status is that we have to assume good faith, especially with book refs that can't be easily checked. It is a problem, though. - Peregrine Fisher (talk) (contribs) 04:34, 17 December 2008 (UTC)

Actually enforce WP:V proposal[edit]

The problem here is that a lot of people remove material that is actually sourced. Not by a ref at the end of a word or a sentence, but at the end of a paragraph containing the fact. And removal of such content makes it easier for vandals. First remove the reference. Then let it sit for a while and contest the content. To make this work properly we need some verification the material is actually unsourced. Another problem is that there are a lot of people who say that a lot of people who don't reference their content nevertheless provide material that turns out to be good when cleaned up. They don't want to lose material because the newbie who made it doesn't know how to reference. So we should also make referencing with the right amount of detail (ISBN, dates and the like a lot easier). - Mgm|(talk) 12:54, 17 December 2008 (UTC)

I agree that we should make referencing easier/more encouraged, I'd be all for some sort of system to warn people when they're adding large chunks of unreferenced text, and help them include a reference. But this suggests a Wikipedia that takes sources more seriously, and is actually willing to remove challenged, unsourced content. Currently, even if you're making a perfectly good-faith challenge of unsourced content, it might get restored and admins might back up that source-less restoration... it's hard to be serious about sources when that kind of stuff happens. Right now, unsourced content can stay in Wikipedia... even in many cases if it's challenged. There's no motivation for people to provide sources, so they usually don't.
As for your objection, I think you can always construct a hypothetical situation like you did, but ultimately, a source could still be presented, and the person challenging the content, if they're challenging it in good faith, should be satisfied. I don't really think vandals running around removing references are a serious problem anyway... even if they were, we can always view article histories. --Rividian (talk) 14:09, 17 December 2008 (UTC)
  • That's why I prefer moving challenged content to the talk page rather than removing it. - Mgm|(talk) 12:13, 18 December 2008 (UTC)
    • That works too. But just adding {{fact}} tags but keeping dubious claims in articles is really why we have this problem in the first place... challenged content should be removed from the article until a source is found, although it can be put on the talk page. If that's our attitude every time one of these conflicts comes up, people will take sourcing more seriously. --Rividian (talk) 18:38, 18 December 2008 (UTC)
Do you quite realize the difficulty of checking every fact tag and every unsourced article and make a serious effort to find sources first on the web and then in print? and then to check and fix all the millions of things tht appear uncontentious but are not in fact accuraely sourced? (And, every two or three years it has to be repeated, because developments happen and knowledge improves. Print encyclopedias have always done that.) The problem is giving people an incentive to do the work--and a secondary problem is that most of our active contributors are not in fact in a position to make proper sophisticated use of print sources, even if they knew how. DGG (talk) 05:13, 20 December 2008 (UTC)
Searching Google books/scholar isn't rocket science, anyone who can use the internet can do it, and usually find a source in a few minutes if one exists. I'm saying that we shouldn't even have tens of thousands of fact tags, people should just make a reasonable look around for sources, then remove the claim if a source can't be found... no longer should we feel obligated to host rumors and nonsense for years because it's "difficult" to look for sources. If someone thinks the claim belongs (because it can be sourced) they can add it back with the source. When we take the attitude that sourcing is just too hard, that it's okay to leave tens of thousands of fact tags in for years, and that people are free to re-add unsourced content even if it's challenged, then it's no wonder that nothing gets sourced on request - there's no real incentive. If we want sourcing, we need to get serious about actually insisting on sources when a claim is challenged. As is, there's always some admin saying "Oh, well... sourcing is actually optional... it's too hard for people to use Google books" so why should we expect sourcing to ever happen? --Rividian (talk) 21:11, 22 December 2008 (UTC)
I would go for enforceable removal of statements tagged with {{fact}} to the talk page after a maximum of one month (less if there's consensus the claim is unlikely, and of course immediate removal on BLPs as now), and deletion of articles with {{unreferenced}} after the same time frame (again, maximum, less so with consensus). If no one cares enough to do WP:SOFIXIT after one month, then WP:SOGETRIDOFIT. Seraphimblade Talk to me 04:31, 24 December 2008 (UTC)
That could be easily gamed. Say I don't like soccer player articles. I could add a fact tag at the end of each paragraph, or one unreferenced tag, to say five articles a day, and I think I would get about a 90% deletion rate after a few weeks have gone by and our volunteer resources have become taxed. - Peregrine Fisher (talk) (contribs) 04:50, 24 December 2008 (UTC)
Getting rid of unreferenced materials is no more "gaming" than fixing typos is. If they're really that lacking in references, then I guess we're due to lose 90% of them. With sports stubs, I'd honestly not be surprised, a lot of sports "biographies" are in need of listification. It blows my mind that people who see writing things and putting a primary directory-style source as a "reference" see cutting and trimming as some type of "gaming" or something done in bad faith. Editors cut every day, comes with the job. Seraphimblade Talk to me 07:33, 24 December 2008 (UTC)
That argument often crops up, but it shows utterly no good faith towards people who add {{fact}} tags and challenge claims, and is much more of a hypothetical problem than one we actually would see (vandals almost never know how a {{fact}} tag works, 99% of vandals are just bored passers-by). At any rate, when you challenge a claim, you are saying you believe the claim is false, or dubious, or is just so extraordinary it requires a citation of some sort, so there's a good reason to remove the content at some point. If you are removing claims just to blank the content with no good reason, it's vandalism and can be dealt with as such. --Rividian (talk) 15:53, 24 December 2008 (UTC)
I'm not worried about vandals. It would take one prolific editor who doesn't like sports stubs, one who doesn't like pop-culture, one who doesn't like each other category of articles, and now 70% of the pedia is deleted. Plus, delete all the sports stubs, and you'll create people who'll go after other categories in revenge. Anyways, I don't think this would achieve consensus and be implemented. - Peregrine Fisher (talk) (contribs) 17:38, 24 December 2008 (UTC)
I'm talking about letting people remove content when they challenge it for a good reason. This is what WP:V is supposed to let us do already. You just have this hypothetical situation which is more of a scare tactic than something that could actually happen... if someone is removing massive amounts of content they don't even challenge, they're a vandal and would be dealt with as such, even if they're a "prolific editor". I'm not proposing we let people remove any claim that doesn't have a citation at the end of the sentence... but rather, that when someone says "Whoa, I think this is wrong, only re-add this claim when you find a source" we don't have admins say "Well, it can stay in indefinitely... finding a source is too hard, just add a fact tag and wait forever". --Rividian (talk) 19:25, 24 December 2008 (UTC)
I don't disagree with removing uncited material that is challenged for a good reason. I just don't think we should set up any automatic deletion of articles. - Peregrine Fisher (talk) (contribs) 19:44, 24 December 2008 (UTC)
Probably not, but at the same time, if someone removes content to the talk page in a good-faith challenge of its accuracy or verifiability, we need to respect that a source should be found before restoring it. Right now, people can restore challenged content without providing a source, and often keep it in articles. That we have such a permissive attitude is a major reason few people bother looking for sources. --Rividian (talk) 22:26, 24 December 2008 (UTC)
I don't know how you encourage that, but it would increase referencing. I think the community will slowly move towards removing uncited material, but it will take a while. Improving referencing is probably what can speed it up the most. It's kind of a chicken or egg thing, but I think referencing is first. That will probably happen when the interface is simplified. Six months to a year. - Peregrine Fisher (talk) (contribs) 22:45, 24 December 2008 (UTC)

WP:V says (at the moment): Any material lacking a reliable source may be removed, but editors might object if you remove material without giving them sufficient time to provide references, and it has always been good practice, and expected behaviour of Wikipedia editors (in line with our editing policy), to make reasonable efforts to find sources onesself that support such material, and cite them.
So "actually enforcing" WP:V, as stated above, appears to mean changing WP:V to remove the phrases about "giving sufficient time" and "make reasonable efforts to find sources onesself". I suspect that will be next to impossible, and suggest, instead, that the policy be changed to clearly distinguish between challenged statements (via {{fact}} and similar) and unsourced but unchallenged statements. For the former, perhaps a consensus could be achieved that such challenged statements can be automatically removed in (say) two weeks (or 30 days or whatever) and that restoring them without adding a source is a violation of WP:V. And perhaps a bot could implement the "automatically removed" part of the policy. Using a bot would be easy if the current format of text {{fact}} were changed to {{fact|text}}). -- John Broughton (♫♫) 01:03, 28 December 2008 (UTC)
I think we should be working toward adding sources to existing material rather than removing material. Much work was done in the early years of Wikipedia of building the encyclopedia when there wasn't the interest in citing as there is now, and much unsourced but valuable information is contained here. It would be inappropriate to destroy that material because people haven't had the time to go back and source their research and other people are apparently unwilling to. Let us work toward citing the material we have rather than removing it. SilkTork *YES! 20:30, 3 January 2009 (UTC)
I agree with SilkTork; however, I find that making a fact request does help. Often when reading an article and I see a fact request, I think, "I'll bet I find that in less than five minutes." and frequently I do. Similarly, where I suspect that the proposition is unsupportable, I have inserted fact requests, and returned to them several months later either to find it cited, removed, or in some cases, I feel no remorse in then removing the suspect sentence. Lets not delete whole hog. If the policy suggests that, then it needs rewriting. --Bejnar (talk) 04:35, 5 January 2009 (UTC)


Simply adding some references doesn't mean much. I've seen enough unsupported statements with a footnote pointing to a source that either did not support the statement at all or it was cherry picked fact that ignored the main body of evidence in the source. So the effort should be to have some confirmation that the article has been reviewed and found actual. Currently the assumption is that because an article is viewed a lot it must be fact checked. This doesn't hold for non-obvious facts. There are science articles with 100K views/month and 1-2 active editors at the most (take Creatine for instance).

The B/GA articles are supposed to be fact checked in the entirety by someone active on the wiki. So, let's give these more exposure. I think this will encourage fact checking. It should be straightforward to transclude Wikipedia:Good_articles/recent to the main page. I would change the wording for the non-wiki-speakers to say "recently reviewed articles" (you can help!) or something like that.

Related-proposal: unify the peer/B/GA review processes[edit]

There are too many ways to do essentially the same thing: review an article. Instead of three processes, a process with three outcomes should be simpler: fail (article remains start class), B (article is verifiable, but does not have enough coverage for GA and/or prose problems), GA.

Discussion below[edit]

  • While it's a good idea, it has an inherent problem. People already have to wait ages to get a GA review to be fulfilled. Swarming the process with more nominations simply means it won't scale and might actually give the opposite effect from what we really want. - Mgm|(talk) 10:40, 15 January 2009 (UTC)
  • On their cite the WebCite people suggest that one possible intern job would be to create a bot who archives references and adds the relevant information to Wikipedia. Someone should really take that on.- Mgm|(talk) 11:01, 21 January 2009 (UTC)

Ban the fact tag[edit]

The {{fact}} tag should be deleted. If you need to add it, you should instead be removing the uncited statement. Stifle (talk) 20:31, 17 January 2009 (UTC)

Not a great idea:
  • Many articles on fundamental topics were written prior to the current referencing standards. There may be references at the end that do cover the article. Most of the current FA-class stuff is actually on more obscure topics that were written later.
  • Sometimes an article starts with a single reference so there's little point in using any inline citations in this case. Then a 2nd person adds some detail, with an inline ref, and suddenly 90% of the article seems questionable.
  • Editors would overuse inline citations even more. Take a look at Building 101 for an example.
  • A fact in organic chemistry synthesis or algebraic topology (to pick some examples) may be evident to educated readers but not the rest. Having lay editors simply delete contents would drive even more experts away. (Take a look at User:V8rik history for a recent example)
  • It would increase edit warring.
So, methinks. Xasodfuih (talk) 01:09, 18 January 2009 (UTC)

Form project[edit]

I believe this should be formed into a WikiProject. Stifle (talk) 20:33, 17 January 2009 (UTC)

Wikipedia:Bots/Requests for approval/WebCiteBOT[edit]

Per a suggestion above, I asked if someone could write a bot to use to combat dead links. Join the discussion if you're interested. - Peregrine Fisher (talk) (contribs) 07:48, 24 January 2009 (UTC)

Why are we so concerned with keeping unreferenced articles?[edit]

If an article cannot be referenced or its creator -despite all the warnings on the screen when you edit- doesn't want to provide edits, why does the community care for it as though it were Holy Writ? If it's notable, someone else - who does care enough to- will create and source it. If it's vanity, spam, or otherwise nn, the sooner gone the better. WP is an encyclopedia; there's supposed to be no WP:OR here. Without references, we cannot discount OR. Let's just axe it and keep the articles for which editors are willing to provide sources. If no refs=quick deletion, the editing community will be submitting better, referenced articles. We were all newbies once, and had it been made clear to each of us (as it now is better on the edit page) that w/o references, your article will be deleted, we'd all have matured faster. The fact that we don't follow through on our own policies and warnings is why we have lots of unreferenced articles. Which, contrary to some, I think is bad. Unreferenced does not equal wrong just as referenced does not equal right; but, no one can easily check (see WP:V). There really ought to be a presumption to not keep unreferenced material - a sentence, claim, paragraph, section, or whole article. A speedy category, that would place the article in user space for some period for an attempt at referencing would be a good idea. Carlossuarez46 (talk) 00:08, 27 January 2009 (UTC)

There are a few reasons why articles that should be kept do have references. A million or two articles were created before referencing was standard. You'd have to "axe" probably 2 million right now. Another reason is that adding references is basically a form of computer programming. Most people don't know how to do it. People in the know can barely keep up with AfD now, plus deleting unreferenced articles would drive editors away faster than they're already leaving, which is pretty fast lately. - Peregrine Fisher (talk) (contribs) 00:16, 27 January 2009 (UTC)
There is also a content divide disparity that if a news event happens tomorrow it will be well covered and there will be little issue to getting reliable sources. Roll the clock backa few years, OK maybe ten, and that starts to drop off. Perfectly notable subjects were written about but it takes more work to find reliable sources that may not be available online. Historical documents, as well, may not be searched online, that is, a picture of the print is there but you have to read and retype any text. These are barriers for many users. Technology will hopefully bridge these chasms but we need to be open that reliable sources may exist but might not be readily available. -- Banjeboi 01:26, 27 January 2009 (UTC)
If someone is an earnest contributor and knows enough about a topic to create an article on it - a little note on their talk page prodding them to add sources, with which they may have unusual familiarity or access is warranted. If it's a one-time editor, and they don't come back to see their talk page, deleting their article hasn't driven them anywhere they aren't already. If they stick around, a message on their talk page asking for sources should bring about the desired effect: sources! Carlossuarez46 (talk) 00:40, 28 January 2009 (UTC)
Why don't you test the theory on a few articles. Say 10. If it works on even 3 or them, I would say we could expand the operation. - Peregrine Fisher (talk) (contribs) 00:55, 28 January 2009 (UTC)
I will over the next week or so. I have been focusing on recent changes and short pages and there's plenty to pick from. :-) Carlossuarez46 (talk) 02:15, 20 February 2009 (UTC)
Sounds good. It would be cool if you post which articles and editors are involved here so we can watch it. - Peregrine Fisher (talk) (contribs) 02:17, 20 February 2009 (UTC)

(outdent):I picked a few randomly among articles flagged this month beginning with "A".

Many articles flagged this month were actually created years ago, and some actually had sources, and some should have been speedied instead of flagged, but these are a good few to work on. Carlossuarez46 (talk) 02:26, 25 February 2009 (UTC)
Should be interesting. - Peregrine Fisher (talk) (contribs) 02:36, 25 February 2009 (UTC)

I said, Do, We, Have, A, Problem?[edit]

This project made me wonder how were doing as a project with referencing, so many moons ago I asked the bot people if they could provide some stats. At User_talk:Peregrine_Fisher#Referencing_stats we finally got some. I would say not bad, but judge for yourself. - Peregrine Fisher (talk) (contribs) 04:56, 10 April 2009 (UTC)

Target of link in previous comment now at User talk:Peregrine Fisher/Archive 15#Referencing stats. - dcljr (talk) 02:48, 12 January 2016 (UTC)

tutorial on quickly finding high-quality sources[edit]

Currently, this Wikipedia:Improving referencing efforts article suggests: "When faced with unreferenced information, an editor should make an effort to find sources before deleting material."

Several people have commented that a person should be able to find a reference (if one exists) in less than 60 seconds. I suspect those people know a few tricks I don't know. And I suspect I'm not the only one -- as Peregrine Fisher pointed out, "adding references ... Most people don't know how to do it." Today one such person asked me "where do you find a source ... ?" Is there a tutorial on quickly finding high-quality sources? Something analogous to the Wikipedia:Finding images tutorial ? If not, could we somehow make such a tutorial? -- (talk) 19:17, 28 July 2009 (UTC)

You're right there should be a tutorial, that's a good idea. Were you aware of Template:Findsources btw? -- œ 23:17, 30 July 2009 (UTC)
I did a quick look-around and found Wikipedia:Referencing for beginners. Not sure if it's what you're looking for though. -- œ 23:28, 30 July 2009 (UTC)
That's a start but that helps after the reference has been found. If there is no other resources available feel free to start up a section at Wikipedia:Article Rescue Squadron/How to or link it the page once it's up and going. I think this would be quite helpful. Wikipedia:WikiProject Resource Exchange may also have a lead on this. -- Banjeboi 11:08, 31 July 2009 (UTC)

Merge to WikiProject Unreferenced articles[edit]

I think this could be merged into the Unreferenced articles WikiProject as a taskforce. Mattg82 (talk) 16:58, 5 March 2010 (UTC)

Removed tag merge not required. Little purpose in doing a merge. Mattg82 (talk) 14:38, 9 March 2010 (UTC)
These pages, though slightly outdated, still have much useful information newbies can learn from. -- œ 10:51, 16 March 2010 (UTC)