Wikipedia talk:Reliable sources

From Wikipedia, the free encyclopedia

To discuss reliability of specific sources, please go to Wikipedia:Reliable sources/Noticeboard.

	This is the talk page for discussing improvements to the Reliable sources page.
	Put new text under old text. Click here to start a new topic. Please sign and date your posts by typing four tildes (~~~~). New to Wikipedia? Welcome! Ask questions, get answers.	Be polite Assume good faith Avoid personal attacks Be welcoming

Archives

Contents 1 WP:MOS subguideline, anyone? 2 Examiner.com 3 Template:Cite newsgroup 4 Circularity issue 5 Electronic Intifada 6 Rulings in an american courts 7 Consensus 8 The Michael Jackson situation 9 Foreign language sources 10 Judging reliability of editing in newspapers and other sources 11 Coverage by citation indexes 12 The Washington Post 13 RfC on reliable sources for Eurovision Articles 14 When a reliable source clearly is not

[edit] WP:MOS subguideline, anyone?

Imo, this guideline could be far more usefully handled and improved if it where a dedicated WP:MOS subguideline. The applicable policy (WP:V) appropriately handles the required minimum threshold, while this page could explain various scenarious in greater detail than a policy, including the ideal case, or how to proceed in the many suboptimal cases where high quality sources are not easily available etcpp. User:Dorftrottel 14:52, February 15, 2008

[edit] Examiner.com

I'm going to have to put out a motion that Examiner.com not be allowed as a viable source for linking to articles. There are a number of Wikipedia articles that cite Examiner.com, and there's a bit of problem with this because the Examiner.com model is based on paying writers for page views. Anyone can sign up to be a writer there, put down some information, and then get paid for getting mass amounts of page views. This, in my opinion, creates great conflict with the idea of NPOV. Rustydangerfield (talk) 05:30, 12 June 2009 (UTC)

I agree. Also, examiner.com lacks most of the safeguards of a reputable news source and many of the writings are the equivalent to blogs. There is some great info in these but it is typically available in other places and typically should be cross checked.Cptnono (talk) 07:52, 18 June 2009 (UTC)

I would support an XlinkBot, a gentler process that bot-reverts sources often added by inexperienced users but I wouldn't support a full blacklist of the source. I have seen some articles that were pretty good and I wouldnt want to discard it entirely. Squidfryerchef (talk) 15:19, 18 June 2009 (UTC)

[edit] Template:Cite newsgroup

It's my understanding that we should, under no circumstances, cite a newsgroup. Then why do we have a newsgroup citation template that's used on at least 100 articles? Ten Pound Hammer and his otters • ^{(Many otters • One hammer • HELP)} 11:39, 17 May 2009 (UTC)

Probably because in some very tightly-controlled circumstances it may be appropriate per WP:SPS, as long as it has been demonstrated in a more bona fide reliable source that the post author is legit. I suppose it's not a lot different to citing, for example, a writer's personal blog for minor or uncontroversial information about the writer or his/her work. Use sparingly! Steve ^{T • C} 11:53, 17 May 2009 (UTC)

In particular, to cite a quotation it is preferable to do so directly. Mangoe (talk) 16:31, 18 May 2009 (UTC)

Newsgroup posts can also be intrinsically reliable sources for certain statements about newsgroups. For instance, alt.religion.scientology notes the posting of a specific rmgroup message aimed at a.r.s - it's hard to conceive of a better cite for that than the rmgroup message itself. --GenericBob (talk) 13:55, 20 May 2009 (UTC)

When it comes to quoting what is said in a source, there is nothing more reliable than the source itself. Whether the source should be quoted at all is another matter. Blueboar (talk) 14:07, 20 May 2009 (UTC)

Sure, notability is often an issue for Usenet-related content (cue discussions about coverage bias), but I'd argue that this particular example makes the cut. --GenericBob (talk) 01:30, 21 May 2009 (UTC)

Usenet forums vary widely in terms of reliability. There are a few that are both moderated by, and mainly populated by, experts in their field. comp.dcom.telecom comes to mind as a professional mailing list. Another situation where a newsgroup citation may be appropriate is a bona fide FAQ posted by someone knowledgeable in their field and revised with input from peers. Squidfryerchef (talk) 15:13, 18 June 2009 (UTC)

[edit] Circularity issue

A source currently being used at Polish-Lithuanian Commonwealth - a featured article - contains text that looks to have been copied verbatim from WP itself. The WP content is older, giving rise to the question of the source's reliability. An editor has suggested that the inclusion of the WP material (not credited in the paper) constitutes, in effect, a peer review of the WP content, and thereby makes the paper usable as a source, and has also suggested that the topic be brought here. See [1]. Novickas (talk) 20:57, 3 June 2009 (UTC)

Wikipedia and Wikipedia mirrors are NOT reliable sources, please see WP:Verifiability#Wikipedia_and_sources_that_mirror_or_source_information_from_Wikipedia Jezhotwells (talk) 22:41, 3 June 2009 (UTC)

Yes, Wikipedia mirrors are not reliable sources, but that is not the question. The question is whether an otherwise reliable source that uses a few passages from Wikipedia is still reliable. I would say that if the use of Wikipedia passages is the only issue, it is still reliable. However, in the case being discussed on the noticeboard, it is alleged that the supposedly reliable source didn't cite Wikipedia. Any source that fails to follow the citation norms for the relevant field is suspect. --Jc3s5h (talk) 00:06, 4 June 2009 (UTC)

I have a problem calling any source that copies wording from Wikipedia (especially one which does not have the intelectual honesty to cite wikipedia for that material) reliable. It definitely should raise red flags. While it might be possible to isolate which material came from Wikipedia and which came from elsewhere, I have no confidence that this non-wikipedia derived material is at all reliable. Blueboar (talk) 02:20, 4 June 2009 (UTC)

I would wonder about a source that tries to pass off Wikipedia as a reliable source, just as I would wonder about a source that tried to pass a supermarket tabloid as a reliable source. However, if a source acknowledged the reliability issues of Wikipedia, and explained why the passage from Wikipedia was sound in some specific instance, I would consider the source to be untarnished. --Jc3s5h (talk) 03:05, 4 June 2009 (UTC)

Is this question for the rs noticeboard? [2]

--Ihaveabutt (talk) 04:52, 7 June 2009 (UTC)

the standard for a RS is editorial judgement. If an editor of responsible press or a responsible academic author decides that WP offers a reliable treatment of something, his work counts a a RS. Some of the best academic authors I know use WP for definitions in computer science. In fact, seeing that is what led me to take WP seriously in the first place 3 years ago. DGG (talk) 01:13, 13 June 2009 (UTC)

If you're serious about this, then you might want to change Wikipedia:Verifiability, which currently reads "In addition, sources that present information known to originate from Wikipedia should not be used for that information, as this may create circular sourcing." Also - the source in question was written by students, doesn't say what level - under the supervision of tutors. See pages 10-11. [3]. Novickas (talk) 00:01, 23 June 2009 (UTC)

I agree. But here we are talking about the case when an author plagiarised from Wikipedia without any attribution. In other words, there is reason to suppose that the author is not a responsible academic. I am not saying this is the case here, since it seems entirely possible that there was once an earlier draft of this book chapter on the web which was copied into Wikipedia. But the practical implications in both cases are very similar. --Hans Adler (talk) 02:21, 13 June 2009 (UTC)

[edit] Electronic Intifada

Is the Electronic Intifada a reliable source? I don't see why it is not, but user:ShamWow has removed it from Haneen Zoubi's article as not reliable, citing here. Thoughts?--TM 18:10, 16 June 2009 (UTC)

This inquiry belongs at Wikipedia:Reliable sources/Noticeboard and/or at Wikipedia:No original research/Noticeboard. -- Boracay Bill (talk) 23:11, 17 June 2009 (UTC)

[edit] Rulings in an american courts

I cant find the policy on court documents. Surely the are considered more reliable than newspapers

--Cogvoid (talk) 03:17, 21 June 2009 (UTC)

Depends. They're reliable as a record of what was said in court, but a good-quality newspaper article is often a better source for plain-English interpretation of what a judgement etc. actually means, which is often what we need for WP - this is a 'primary source' issue. --GenericBob (talk) 06:10, 21 June 2009 (UTC)

[edit] Consensus

"The existence of a consensus within an academic community may be indicated, for example, by independent secondary or tertiary sources that come to the same conclusion. The statement that all or most scientists, scholars, or ministers hold a certain view requires a reliable source. Without it, opinions should be identified as those of particular, named sources. Editors should avoid original research especially with regard to making blanket statements based on novel syntheses of disparate material."

This guideline seems to contradict itself, depending on how you read it. I've always taken it to mean that you should not report a scientific or academic consensus unless someone else (in a reliable source) has explicitly said that one exists. Is this the intent? I think getting rid of the first sentence would fix it, but then that seems to leave a loophole where an otherwise reliable source makes a left field claim for a consensus that doesn't really exist. I'd like to get it clarified one way or another, so if anyone has any suggestions, lets have it. Gigs (talk) 19:36, 24 June 2009 (UTC)

Since there's no comment so far, I'm going to be bold and attempt to rework this section. I'm extremely open to discussing the changes I'm about to make if someone has a problem with them. Gigs (talk) 01:43, 26 June 2009 (UTC)

I would advise otherwise. This will have some significant implications for some articles. The consensus does need to be shown if challenged. This can be done either by explicit comment or by implication--ie use of a standard textbook. Personally, I find that many statements in the published literature from a party on one side of a controversy about what the scientific consensus is are somewhat biased, typically being examples of denying the presence of an opposition. The sttement of even a sound atuhority on the subject must still be attributed, and if different statements of consensus can be found, they must be included also. DGG (talk) 23:18, 26 June 2009 (UTC)

I think you and I aren't too far off. The old wording seemed to imply that a Wikipedia editor could look at various sources and then declare a scientific consensus. I want to get rid of that implication, since that's almost the definition of synthesis. Gigs (talk) 01:58, 30 June 2009 (UTC)

[edit] The Michael Jackson situation

Those interested in policing/updating/quality controlling WP:RS and related policies should take note of the issues occurring in the Michael Jackson article today (note particularly the talk page) in which undeniably reliable sources with editorial oversight -- the Associated Press, the BBC, CNN, etc. -- were being rejected outright as RS with regards to reporting Jackson's death. No less than CNET has even posted a story regarding this controversy here. I can understand the reluctance to go with information originating from the gossip site TMZ, but things got out of hand when the Associated Press was being rejected. I think the conduct regarding RS has set a bad precedent. I'm not pointing a finger at any user on this issue, but it is worth investigating for future updates to the policy. 68.146.81.123 (talk) 23:26, 25 June 2009 (UTC)

I think you miss the point. When the reliable sources were reporting that there are reports of Jackson's death, these were quite rightly interpreted as not confirming the death. When RS began to report the death as confimed by the LA Coroner, they were accepted. Jezhotwells (talk) 08:48, 26 June 2009 (UTC)

For a while there, these news outlets were reporting that TMZ was reporting that MJ was dead. This, of course, is not the same thing as news outlets reporting that MJ was dead. I think we did fine. A Quest For Knowledge (talk) 12:49, 26 June 2009 (UTC)

(speedforceironman1809081389099 00:53, 27 June 2009 (UTC))

[edit] Foreign language sources

Is it appropriate to cite sources in a language other than the one the Wikipedia entry is written in? The problem that arises is that foreign language sources necessarily are less comprehensible by possibly a majority of other contributors. However, in my view, sources written in those languages that are most widespread, like English, Spanish, French or German, should be accepted for citation. Sources written in such languages might easily be verified by other users. --Hcinmmod (talk) 19:57, 27 June 2009 (UTC)

See WP:NONENG. Wtmitchell (talk) (earlier Boracay Bill) 00:02, 29 June 2009 (UTC)

[edit] Judging reliability of editing in newspapers and other sources

How are we judging the comparative quality of newspapers? They all have editors, but they are all obviously not equal.

Tabloids like the Natioinal Enqurier are obviously of inferior quality, but what method are we using to judge this? First, you need some method to determine if something is a Tabloid or simply reports on celebrities a lot (like People). Second, if it isn't a tabloid, you still need to employ some other methods to judge quality as a reliable source.

I think that people are frequently applying personal opinion, for better or for worse, as to what newspapers are more reliable than others, as it's not like there's some database of "editor reliability" and "editor expertise."

The only thing I can think of that's objective is circulation of the newspaper and how often the journalist/editor gets printed, which is an indirect measure of the consensus among journalists and their readers regarding the writing subject.

This is a universal method, because even with scientific journals, we judge how often a scientist gets published, is cited, and accepted within the scientific community.

If we are using circulation (in essence, popularity as a measure of consensus), why can't the same be applied to non-expert websites and blogs? It doesn't make sense to allow popularity of a source as viewed as being reliable for one type of source, but not others.

In other words, if Joe's website is frequently linked to by others as a good source to read (i.e. linking to him because they think the site is good, not bad), it would make sense to use Joe as a reliable source. You'd compare the proportion of recognition from all sources.

The quality as a reliable source would be weighted against other factors, such as whether there were multiple contributors, whether it's equivalent to an "op ed" piece or written in a more professional fashion, what sources they used for their information, etc...

Thus, I think that proportion of recognition as a reliable source (recognition within the relevant population e.g. the scientist population for journal articles) should one of the main indicators for all types of sources. It's not the only indicator, but it's an important one.

-Nathan J. Yoder (talk) 16:02, 29 June 2009 (UTC)

I don't think that that circulation is a good way to determine a source's reliablity. Many people read a publication for entertainment value rather than for knowledge. Cosmopolitan and Maxim have circulations of 2,928,041 and 2,558,475 whereas higly respected academic journals such as Nature have a circulation of only about 65,000.

Ultimately, it boils down to whether a publication has a reputation for accuracy and fact-checking. One way to do this is to look at the number of awards it has won. The New York Times, for example, has won 101 Pulitzer Prizes. Another indication is to look at how many times other reliable sources cite it. For academic journals, they actually have a Science Citation Index and Social Sciences Citation Index. A Quest For Knowledge (talk) 16:54, 29 June 2009 (UTC)

You're comparing newspapers to scientific journals, which I clarified are assessed via different populations. As I said, the relevant population for scientific articles is the population of scientists, so you'd do what you just said--use citations and positive views from within the community. Plus, it's not just subscribers within the community, but anyone who regularly reads their articles and a lot of scientists read Nature.

There isn't really a readily available source for how much a print article is cited, beyond perhaps using Google or Alexa to see how much it's linked (if it's available online). In those cases, I don't see how that's different than judging quality of a random webpage the same way.

How do you determine which awards are "acceptable"? There are a variety of them from different sources, many with political and ideological leanings. I don't think a select, "elite few" should determine, for the rest of the world, what's the best. It's largely subjective (i.e. their authority is largely arbitrary) and expertise doesn't really apply in judging here (unless they are an expert in the specific area the newspaper article covers e.g. a geologist critiquing a journalist's geology article), which is why reliability should be left to the whole population of people who read that type of publication.

Besides, Pulitzer Prizes are usually awarded to the journalists, not the newspapers, and journalists often write for multiple newspapers, so then you've got a counting issue. There's also the fact that very few journalists have won such a prestigious prize; you're talking some small fraction of journalists, which would require limiting yourself to a very small number of articles (they award one prize per category per year).

People almost never win them twice and there's no assurance that past prizes hold the same value as they did back when it was originally awarded. If a specific journalist is covering something in an area related to where they won some prestigious award or recognition, that adds value to their reliability, but it's rarely the case that you'll get an news/magazine article like that.

In this day and age, the number of journalists and print publications is rapidly growing and many are reliable, non-tabloid sources, but comparatively little will win a prestigious award and that percentage will only decrease with time, because the number of new journalists per year easily over shadows the few awards given per year. -Nathan J. Yoder (talk) 17:31, 29 June 2009 (UTC)

You bring up a lot of points, and I don't have time to answer all of them but it still boils down to whether a publication has a reputation for accuracy and fact-checking. You determine this by seeing what other WP:RS say about it. Same thing with awards. Whether it's a Pulitzer or the Buckeye News Hawk Award, you look it up. Other indicators that we use to determine reliability are really secondary. For web sites, we often check to see if there's an About page, an editorial staff, named and professional writers, physical address, etc. but these are ultimately secondary to the core issue: a reputation for accuracy and fact-checking. You'll note that Snopes.com fails many of these tests, but it is considered a reliable because it has earned an excellent reputation for accuracy and fact-checking. A Quest For Knowledge (talk) 18:48, 29 June 2009 (UTC)

[edit] Coverage by citation indexes

I modified an earlier change to indicate that not all academic fields are covered adequately by such indexes. I am open to other wording, because I think the problem here is a little complicated. For one thing, the competition between Scopus and Web of Science has caused them both to increase coverage, which is good for indexing, but somewhat decreases the quality distinction. Then, Google Scholar is among other things a citation index, and we do not mean to include it here. Third, it's not covered/uncovered--the coverage in, say , biochemistry , is not as complete as in mathematics where there are more small journals. Fourth, the distinction can not deal with the problem of journals of high quality from the first. fifth, the less developed countries are poorly represented (China though has one of its own--does anyone know about Japan?). The general principle that a source is reliable if qualified people think it is, is certainly true, but inherently a little circular. . DGG (talk) 02:35, 30 June 2009 (UTC)

You can't determine that based on what other reliable sources say about it, because then you're stuck in circular justification, because you then have to demonstrate that the sources you're using as evidence of how reliable a source is are reliable, themselves. We're left with awards, but like I said, that creates their own major problems, which you haven't yet addressed (the rarity of them, subjectivity, and lack of a way to determine what awards are reliable). What does a physical address have to do with the quality of a website? How are you determining if they are "professional writers? Snopes.com fails ALL of those. It has only two writers (not professional) and no prizes. Snopes is considered reliable by web users and we know that based on site usage statistics, which is the criteria I proposed before.

DGG, I think you misunderstood the purpose of a citation index. A good citation index will tell you how many times an article is cited and even how much an author's articles as a whole are cited, as well as other citation measures. The more they index, the better the data. It's about how much scientists cite other scientists, not how much they've been indexed in search engines. They're usually pretty complete. You could also check other media sources that cite certain journal articles. -Nathan J. Yoder (talk) 06:00, 30 June 2009 (UTC)

Sorry for some confusion. I was not talking about web sites, anyway, or about snopes. I forgot to put in a heading, so it looked like my discussion followed on to that one. I've put it in now.

I think I know very well the purpose -- and limitations-- of a citation index, having seen and used them since the very beginning of SCI, and having taught their use at 3 universities. And having written a multi-part peer-reviewed comparative review on them-[4], [5], [6]. There's also 2 nice papers by a student of mine [7], [8].

Incidentally, i fully support the employment of site usage statistics to determine notability of a web site. It does not however determine reliability, or WP would be a RS. DGG (talk) 23:09, 1 July 2009 (UTC)

[edit] The Washington Post

I do not belive that The Washington Post is an excellent example of a reliable source because of its lack of political neutrality. Instead, may I suggest a newspaper which does not have a political leaning such as the Wall Street Journal. I think this website needs to make sure that all articles are neutrally written and neutral sources are cited. Often times i find articles severely edited towards the "liberal" side, and no one seems to revert the edits. —Preceding unsigned comment added by Bonked116 (talk • contribs) 02:06, 7 July 2009 (UTC)

"Instead, may I suggest a newspaper which does not have a political leaning such as the Wall Street Journal." Unfortunately, I was drinking tea when I read this. Now I have to clean my computer screen. --Jc3s5h (talk) 02:51, 7 July 2009 (UTC)

[edit] RfC on reliable sources for Eurovision Articles

I should probably have posted this here before. RfC on the reliability of sources such as ESCToday, oikiotimes is open at Wikipedia talk:WikiProject Eurovision#RfC on reliable sources for Eurovision articles. Thanks. Jezhotwells (talk) 09:21, 10 July 2009 (UTC)

[edit] When a reliable source clearly is not

I originally tried to post on this topic in RS/N but it got hijacked by a specific case and didn't attract disinterested commentary, so it might be best discussed here in the broad sense. I reasonably regularly come across ostensibly RS/V sources, for example articles in journals or newspaper reports, however the articles have obvious, glaring factual errors (usually without citation) that are easily disproved through better sources. In journals these usually occur when discussing cross-discipline topics, such as a pyschology journal discussing business. In newsmedia, well they occur all over the place! How should these types of sources be handled? Someone wanting to use them to support some fact can clearly claim they're from an ostensibly RS/V source and want to use it on WP. On the other hand, clear evidence the article is not reliable can not be included in WP as it would be OR. The current guidelines don't seem to address this problem. --Insider201283 (talk) 14:02, 10 July 2009 (UTC)

What you are essentially discussing is comparative reliability... ie determining whether one source more reliable than another. The reason why we don't get into this in the guideline, or at places like RSN, is that such determinations simply can't be mandated by policy... they have to be discussed and determined at the article level, by those who know the subject and the specific sources under discussion. All we can say at a guideline level is that we hope that our articles to be based upon the most reliable sources available. We leave it to our editors to figure out what those sources should be. Blueboar (talk) 14:32, 10 July 2009 (UTC)

Wikipedia allows source-based research, but not original research. If I find one apparently reliable source that says the first human moon landing was in 1996, and 25 reliable sources that say it was in 1969, it is source-based research to write in Wikipedia that it happened in 1969, with no mention of 1996. On the other hand, if a reliable source says an automotive headlight draws 2 MA of current, it would be original research to measure the current drawn in a headlight of my personal vehicle and write in Wikipedia that automovive headlights draw about 2 A of current. --Jc3s5h (talk) 14:55, 10 July 2009 (UTC)

However, we can use original research on the article talk page when disussing the relative merits of a source in comparison to another. We can't put any OR in the article itself, but OR arguments that point out the flaws of a source can help us determine which source might be more reliable than another.

Also... something I forgot to mention... when I say we strive to use the most reliable sources possible, we also have to take into account WP:NPOV... if the two sources are expressing equally significant, but differing points of view on a topic we should mention what both sources have to say on the matter, even if one is considered more reliable than the other. Balancing POVs is important (and not always easy). Blueboar (talk) 15:06, 10 July 2009 (UTC)

I understand that with regard using the source for a specific fact or point. My question is more when someone wants to use a source to support claim A, and there are few or no sources to either support or oppose the assertion - but elsewhere in the source there is also assertions B,C,D - and there are many sources showing the source is wrong on B,C,D. It would seem to me that this would question the reliability of using the source for claim A. So if you read an article that says "there's a tea cup orbiting the moon", and you can't find anything that says its wrong, but elsewhere the article also states "the moon is made of green cheese", "Neil Armstrong was the first mouse in space" etc - clearly this would weaken claims of reliability for the first assertion. An exaggerated example, but hopefully you get my point. --Insider201283 (talk) 15:17, 10 July 2009 (UTC)

I would agree that it would weaken claims of reliablilty for the first assertion. However, whether it weakens them enough to exclude the source (and the statement) from use in the article has to be determined by debate and consensus on the article's talk page. It isn't something we can determine at a policy level because the issue is specific to the particular source. Blueboar (talk) 15:26, 10 July 2009 (UTC)

Ok, but at present the guidelines don't really speak to this at all. It talks about the author and the publishers contributing (or otherwise), but not the text itself. Perhaps the problem is more that the guidelines as written imply that if it's published in a peer-reviewed journal, then it's ok no question. Now they don't actually say that, but it's easy to understand how editors could interpret it that way. --Insider201283 (talk) 16:03, 10 July 2009 (UTC)

Policy statements have to be based on normal conditions and situations... and you are discussing an abnormal one... an exception to the norm (a source that dispite being in a peer reviewed journal may not be reliable). If you look hard enough, you can probably find exceptions to every policy statement on Wikipedia. These exceptions don't invalidate the policy statements. Most papers that appear in peer reviewed journals are going to be highly reliable sources. Look to the intent of the policy, and less to the narrow wording. Blueboar (talk) 16:47, 10 July 2009 (UTC)

Fair enough. Unfortunately not so abnormal in some of the areas I'm often dealing with.:( So it's off to RS/N to duke it out on a case by case basis I guess. Thanks for your input. --Insider201283 (talk) 17:17, 10 July 2009 (UTC)