Wikipedia talk:Reliable sources

Defining reliability of a medium via the trust on the medium among its readers

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

---

Hi all, I'm new here and writing with background of journalistic studies. Mass media is commonly used as a reliable source in several Wikipedia article and referred as "a reliable source". I'd like to suggest adding the following into the definition of reliable sources:

When citing a medium as a reliable source, there should be an independent study or poll supporting the reliability of that medium. For example a poll showing more than 50% trust among the readers would support the reliability, and in the opposite less than 50% would support unreliability. This should not be taken as a requirement, as such information is not available for all the mediums, but for those who have it's a good rule of thumb when defining the reliability. This rule of thumb should not be used for defining a reliability of a single article published in any medium as any article might be reliable despite the trust on the medium in general. However, reliability of a medium shouldn't be founded solely on trust among readers as there are areas, times and circumstances where/when mediums are used for propaganda or for other purposes. Please also refer to other methods identifying reliable sources. The reliability of a single article should always be handled separately considering the concept, but more focus on reliability should be addressed especially when using articles in mediums having low public trust among their readers. When a reliability on a medium(s) plays important role in a Wikipedia article, and especially when using article(s) of medium(s) carrying low public trust, it's a good practice to reason the use of such medium and when possible, link readers to a recent study or poll on the reliability of the medium.

EDITS IN THE ABOVE SUGGESTION:

1. After comment by Hob Gadling, edited adding However, reliability of a medium shouldn't be founded solely on trust among readers as there are areas, times and circumstances where/when mediums are used for propaganda or for other purposes. Please also refer to other methods identifying reliable sources. Edit by 81.197.179.232 (talk) 11:32, 6 April 2017 (UTC)

DISCUSSION:

I think the above suggestion requires a discussion in advance both on it's usefulness, topic and the content itself. My goal here is to ensure good practices when referring to mediums as a source. Thanks for any opinions. 81.197.179.232 (talk) 13:48, 5 April 2017 (UTC)

I stopped reading after "When citing a medium as a reliable source, there should be an independent study or poll supporting the reliability of that medium." No writer searches for an independent study or poll every time the writer wants to cite a source. This idea is a non-starter. Jc3s5h (talk) 13:59, 5 April 2017 (UTC)

My suggestion does not require writer to do that. Please read through at least, and maybe suggest edits which would remove such an impression you got. 81.197.179.232 (talk) 14:24, 5 April 2017 (UTC)

By the criterion of whether the readers agree with the articles, Völkischer Beobachter and Pravda are reliable sources. --Hob Gadling (talk) 19:45, 5 April 2017 (UTC)

We use polls (at least among ourselves) to establish media reliability only when a dispute arises. Heck, even Pravda may be a reliable source when it says that "Khrushchev visited kolkhoz "Lenin's Glory". Staszek Lem (talk) 21:17, 5 April 2017 (UTC)

Thank you for bringing up this point of view. You refer to mediums in a country under dictatorship / communism, so we should also consider the reliability of such polls. I suggested an independent study or poll which includes this level of reliability check when needed, but it might not be enough (There actually might be trust among readers). An edit to my suggestion is in place. My suggestion is not supposed to drop any other method regarding reliability checks on mediums, like for example how a propaganda medium is currently handled as a source in Wikipedia, and the suggestion should not enable the use of such mediums any easier than the current practices are. Having these in mind, I've edited my suggestion by adding However, reliability of a medium shouldn't be founded solely on trust among readers as there are areas, times and circumstances where/when mediums are used for propaganda or for other purposes. Please also refer to other methods identifying reliable sources. Again thanks a lot, happy if you check if this edit covers the issue (and doesn't bring up any new issues). 81.197.179.232 (talk) 11:32, 6 April 2017 (UTC)

Actually, my two examples were intended not as constructive suggestions for improving the proposition but as counterexamples for destroying it. What you are suggesting is to use the fallacy of argumentum ad populum. People are fallible and easy to fool, in democracies as well as in dictatorships, and most of the time, in all countries, the majority holds several batshit crazy ideas with no connection to reality, often because media spread them. Sometimes the majority elects known liars and frauds into positions of power because of that. Truth and reliability are not, and should not be, determined by vote. --Hob Gadling (talk) 14:30, 6 April 2017 (UTC)

I don't mind if your goal is more destructive or constructive as long as it's well argued, which you again do. My counter argument: As long as gallups, polls or other measures of opinion are used to support scientific studies I can't find a reason why they should be disallowed when determining reliability of a medium. You also shared a valuable point re argumentum ad populum but I see that already taken care of with the latest edit. In the current form the reliability is not based solely on the fact that it is a widely popular thought and it does not disregard the importance of expertise, independence, accuracy and fact-checking, but instead adds to that list. 81.197.179.232 (talk) 18:41, 6 April 2017 (UTC)

English can be a confusing language... It took me a moment to realize that we were (once again) discussing news media, and not discussing the reliability of mediums. For a moment there, I thought the OP was asking us to search for polls to determine whether (for example) Theresa Caputo was more trusted than John Edward... or something like that. Blueboar (talk) 12:42, 6 April 2017 (UTC)

It can... a medium was a generally used term during my studies to cover any intermediate agency or a channel of communication like as a single instance in the field of media, definitely not limited to news. Do you think another term would be better to avoid confusion? 81.197.179.232 (talk) 18:41, 6 April 2017 (UTC)

How long time should we give for the discussion here before considering editing the page? (assuming the suggestion would benefit Wikipedia) 81.197.179.232 (talk) 12:59, 8 April 2017 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Challenges to the closure should be taken to WP:AN (after contacting the closer, which has taken place), not WP:TO, per WP:CLOSE#Challenging other closures – Finnusertop (talk ⋅ contribs) 15:09, 10 April 2017 (UTC)

secondary source cite of primary source

For a case where a primary source x says “We conclude with y degree of certainty that z” where the primary source defines ‘y degree’ as being less than 100%, is a secondary source’ considered reliable when it specifically cites the primary source a statement of the form “<primary source x> concluded y” rather than qualifying that statement as the source did? Related: Are there degrees of reliability that pertain here? Humanengr (talk) 00:48, 18 April 2017 (UTC)

Issues like this frequently arise when one source cites another inaccurately. I don't think it has anything to with primary vs secondary. Zero^talk 00:56, 18 April 2017 (UTC)

Thx. So ignoring the primary vs secondary aspect, would the citing source be considered reliable if it cites 'inaccurately'? The answer seems an obvious 'no'; I'm just confirming re WP use of the term 'reliable'. Humanengr (talk) 01:03, 18 April 2017 (UTC)

Elaborating: RS:News organizations says "'News reporting' from well-established news outlets is generally considered to be reliable for statements of fact (though even the most reputable reporting sometimes contains errors)." Here the citing source is stating that another source "concluded z" without the qualifier. The issue is whether it should be considered 'reliable' when it is not 'accurate'. Humanengr (talk) 01:15, 18 April 2017 (UTC)

The problem is in deciding who gets to decide what is "inaccurate". For example, the usual standard in lots of inferential statistics is 95% confidence that a result is not random. But the results are often described as given - as a conclusion, that yes, something or other is as it is. But sometimes, depending on the research question and the data involved, the standard is 90%. Or 99%. So what's "inaccurate" here?

And here is the thing. Based on experience, I'd say that a lot of Wikipedians probably have some difficulty even with the whole "verifiability" part of sourcing - is it sourced or not? The last thing you want is this bunch running around trying to evaluate whether or not a source is being "accurate" or not. That way lies madness. And reddit. Basically that would just turn Wikipedia into another crappy internet discussion forum.Volunteer Marek (talk) 01:37, 18 April 2017 (UTC)

Re "depending on the research question and the data involved". Does that include consequence (say in terms of benefit or cost) of the decision? Humanengr (talk) 04:57, 18 April 2017 (UTC)

In theory, yes (Loss function). In practice, most often it really comes down to the quality of the data. But that's not a Wikipedia topic.Volunteer Marek (talk) 05:13, 18 April 2017 (UTC)

I'm asking whether you are saying that consequence should or should not be considered. The probability of an event is one thing; the consequence of an event is another. I am not asking about the quality of the data for either, only whether consequence should be considered. Humanengr (talk) 05:39, 18 April 2017 (UTC)

Are you asking about Wikipedia practice specifically, or just scholarly work in general? If latter, then yeah, "consequence of an event" is part of the loss function - look at that article.Volunteer Marek (talk) 18:09, 23 April 2017 (UTC)

Let's pick this up later after discussion below re 'secret' (beyond the Taleb issue). Humanengr (talk) 19:44, 23 April 2017 (UTC)

This is a hard question and I don't think there is a simple answer to it. Some points: (1) If both the sources are independently citable under the rules, you can cite them both to show the disagreement. However, you shouldn't write it as a story of how one source got it wrong (which would be a SYNTH violation). Just write that one source says A and another source says B. (2) Rules like NOR and V don't apply to talk pages, sources can be reliable for some things and not for others, and editors have the right to choose which items in "reliable sources" get into articles anyway. It is perfectly ok to argue on the talk page that a source made a mistake, using whatever evidence you can muster, and if a consensus forms that you are right you have decided that the source is not reliable for that fact. Zero^talk 01:52, 18 April 2017 (UTC)

I'm not sure what "independently citable" means (and don't see a mention of it in the RS article. Re "Just write that one source says A and another source says B.": that would result in, e.g.:

"<Name of 1ary source> concluded that y did z.[cite to 2ary source] <Name of 1ary source> "concluded with high confidence that y did z.[cite to 1ary source]"

Is that what you are recommending?Humanengr (talk) 12:22, 18 April 2017 (UTC)

"Independently citable" just means that each of the sources satisfies our citing rules (like WP:RS) on its own merits. As for how to cite both of them, not every problem has a solution which is both satisfying and elegant. One device you might think about is to put the secondary source's opinion in the main text, then in the footnote giving the full specs of the secondary source you can note that it cited the primary source which you then quote. It is perilously close to SYNTH, but maybe ok if you don't write that the secondary source made a mistake. Zero^talk 13:22, 18 April 2017 (UTC)

@Zero0000: Pls see discussion below with Jc3s5h wrt 2ary sources citing 1ary public summaries of their 'secret' investigations. Comments welcome. Humanengr (talk) 20:33, 23 April 2017 (UTC)

I should also add that in my experience on Wikipedia I've really only seen a few instances where a secondary, reliable, source really "got it wrong". 90 times out of a 100 it was the user who wanted to use primary rather than secondary, who either did not understand the primary source or was pretending not to understand it in order to push their POV. Out of the remaining 10% a good chunk just involved sources which used some ambiguous phrasing which could be easily misconstrued. The couple times where the source "got it wrong" it actually wasn't that hard to get consensus on talk not to use that particular source, or to write it in the way which makes it clear that something fishy is going on (for example, do it the way Zero0000 suggests).

Basically, if you think "source got it wrong" and others don't see it that way on the talk page, there's a pretty good chance that you're the one getting it wrong.Volunteer Marek (talk) 05:13, 18 April 2017 (UTC)

@Volunteer Marek: Pls see my response below to Jc3s5h re a narrower focus; will return to broader issues later. Comments welcome Humanengr (talk) 17:04, 23 April 2017 (UTC)

"We conclude with y degree of certainty that z" is a typical way to state a conclusion in a scientific journal. "Jones concluded z" or even just "z" may be a typical way to state a conclusion in a source aimed at a popular audience, or when addressing a technical audience in a briefer format. So it comes down to whether the author of the secondary source has the qualifications to make the jump from "Jones concluded with y degree of certainty that z" to "Jones concluded z". If the secondary author is a scholar in the field, I'd consider it reliable unless there are other reliable sources contradicting the statement. If it's a general-interest reporter writing a newspaper article about a technical subject, I wouldn't be so sure the secondary source is reliable for the statement about z. Jc3s5h (talk) 14:44, 18 April 2017 (UTC)

@Jc3s5h: Thx. In science, experiments are repeatable (more so in physics, less so in other disciplines), so there is a possibility that a 2ary author -might- have some relevant experience and -might- be qualified. But in, say, writing about a public summary release of a secret investigation by a government agency, the 2ary author has no basis on which to make a judgment of the 1ary source designation of confidence. Any change from the designation given by the 1ary source -- e.g., "with high confidence" is a biased distortion. Humanengr (talk) 19:34, 20 April 2017 (UTC)

I must disagree. Making use of the conclusions of the primary source requires readers to transform the statement to different wording or action. A technical reporter writing for a lay audience writes "Jones concluded z." An engineer writing a test spec translates "a properly functioning circuit will have an output voltage that is normally distributed with mean 1.0 V and σ 0.01 V" to a program that shunts any part with an output voltage >1.03 V or < than 0.97 V into the trash bin. A patient translates a probabilistic statement about the risk of a surgery into a decision to have the surgery. Making use of a statement involving probabilities involves transforming the statement into a useful form for the situation. Jc3s5h (talk) 21:26, 20 April 2017 (UTC)

Agree that 'use' and 'decision' are key. Are you familiar with Taleb? Humanengr (talk) 21:53, 20 April 2017 (UTC)

No. Jc3s5h (talk) 23:17, 20 April 2017 (UTC)

@Jc3s5h: Taleb's proofs of errors in "traditional treatment of probability distributions" have radically changed decision-making but I can leave that aside for now.

Not really. Taleb's audience is mostly the semi-educated who wish to seem smart without putting in the work. He's basically saying the same shit that's been known for years just dressing it up in new phraseology and pretending it's original. The loss function (multiplying consequences by their probabilities and basing decisions on that) has been around for a couple centuries implicitly and explicitly for almost a full century.Volunteer Marek (talk) 18:12, 23 April 2017 (UTC)

(by "semi-educated" I mean people with a Masters or a Bachelors degree, depending on the discipline ;) )Volunteer Marek (talk) 18:18, 23 April 2017 (UTC)

I don't want to pursue this now except to say that's not even wrong. But I'd welcome comments below re 'secret'. Humanengr (talk) 02:07, 24 April 2017 (UTC)

Ok. But if you think that Taleb "proved" that there are errors in "traditional treatments of probability distributions" or that his stuff has "radically changed decision-making" (how would you know?) then, well, that's "not even wrong".Volunteer Marek (talk) 08:58, 25 April 2017 (UTC)

In your engineering and surgery examples, as in my physics, etc., examples, the data and methods have been made publicly available. In those cases, another party -might- have some basis for "transforming the statement into a useful form for the situation". But in the case of a secret report they do not have access to the data and methods and so have zero basis for "transforming the statement". Any change is a distortion. Humanengr (talk) 08:13, 23 April 2017 (UTC)

• Responding to the OP - every policy and guideline says that we should rely on secondary sources, and I will add to that, that we should rely on high quality secondary sources. There are zillions of reasons for this. One reason to reach for high quality secondary sources is that their expert authors will gather up the underlying primary sources and contextualize them. It isn't our place as editors to peer review them. We should just summarize them. One of the ways that editors go wrong is that they grab low quality secondary sources that don't deal well with the material they are discussing; editors also go wrong by citing primary sources and interpreting them in various ways.

With the way the question is phrased, neither I nor anybody else can understand if the secondary source is high quality (a New York Times article, or a review in the NEJM) or low quality.

In general we defer to high quality secondary sources. 20:53, 23 April 2017 (UTC)

@Jytdog: I narrowed the discussion in my post immediately above (which basically repeats my 19:34, 20 April 2017 cmt). Can you respond specifically to the issue of 2ary sources citing 1ary public summaries of their 'secret' investigations? Humanengr (talk) 21:04, 23 April 2017 (UTC)

The question remains too vague to answer. You give no sense of whether the 2nd source is high quality of low. If it is Daily Mail-ish and it mischaracterizes the underlying source of course you don't use that. These kind of hand-wavy discussions are generally unproductive. It would be much better if you actually brought the sources you are concerned about. Jytdog (talk) 21:11, 23 April 2017 (UTC)

btw I looked at your editing history, and am guessing this is something about the Russian interference with US elections and what the US intelligence agencies found, how those findings were reported, and what other people think about those findings and of various reports of them. This is a morass where obvious political agendas are making editing more difficult than usual. Per the advice in WP:Controversial articles, concentrate on raising source quality and push away efforts to introduce content based on low quality sources. Jytdog (talk) 21:15, 23 April 2017 (UTC)

I have no problem in talking about the specifics; but the issue is broader. The issue is what is considered 'high quality'. You are correct that political agendas are of concern, and that is exactly what I am focusing on. I want to address this more generally but will note the particulars here (which include some confounding issues).

In this example, the 1ary source, the ODNI, provided a Jan 2017 public summary of their conclusions as "We have high confidence in these judgments." Obviously, as for all secret investigations, they did not publish their data or methods and so are not available for review and critique. The 2ary source cited for the lead sentence inexplicably refers not to the Jan report but to what it superseded -- a Oct 2016 'joint statement'. This results in a lede of "The United States Intelligence Community officially concluded that …" without the 'high confidence' qualification. The cite to the Jan 2017 report is postponed to the 2nd sentence. The discussion on the talk page has promoted the 'concluded that' language citing various later sources. But none of those later 2ary sources are cited in the lead para. So the controlling presentation to the reader is that the -current- assessment is 'concluded that'.

I hope those details provide context for what prompted my general concern, which I restate here eliminating the confounding issues in the example above: Assume, for sake of simplicity, that a 2ary source cites and rephrases the public summary given by 1ary source of their 'secret' investigation. The 1ary source, irrespective of any supposed stamp of 'quality' approval by the WP community or anywhere else, has no basis -- since the 'secret' nature of the investigation prohibits release of data or methods for critique -- for rephrasing the 1ary's statement of their conclusion -- except a political agenda or incompetence, both of which make the source extremely unreliable for any such statement.

This applies to all 2ary reports of secret investigations. Humanengr (talk) 22:16, 23 April 2017 (UTC)

I'm less convinced than before. The report defined three confidence levels then assigned the highest level to its main conclusion. It isn't a qualification that lowers the strength of the conclusion. It allows for a small possibility of being wrong, but so does "officially concluded" (as opposed to "determined beyond any doubt" for example). So in this case I think the secondary source's summary was not too bad. Zero^talk 02:01, 24 April 2017 (UTC)

Interesting then how all discussants opposed to changing the lede to (in your view) the stronger statement are those convinced the claims are undoubtedly true. Humanengr (talk) 02:34, 24 April 2017 (UTC)

Just FYI - people's strong views make it harder. Guessing at what people's views are and making comments about that in the midst of discussions, make things impossible. If you keep doing that you will end up topic-banned. WP:FOC. Jytdog (talk) 04:11, 24 April 2017 (UTC)

@Zero In your view, does "officially concluded" allow for different interpretations? Humanengr (talk) 09:32, 24 April 2017 (UTC)

The word "officially" can have different import depending on context. On its face it just means that the committee presented the report as a formal action, in which case it has no bearing on the certainty of the conclusions. However, authors sometimes attach the word "officially" to drop a hint that something might be hidden, i.e. that there could be an unofficial position that differs from the official one. I have no idea if that applies here. As for "concluded", that's what the committee did: they wrote "conclusions". All committee conclusions come with a level of certainty, though usually it isn't stated. Not saying that a level of certainty was attached, especially if it was the highest defined level, is not a sin if every reader should know that some level of (un)certainty is necessarily there. If you think that "officially concluded" means "proved" or "determined as a fact", you are simply mistaken; it does not mean that. Zero^talk 10:07, 24 April 2017 (UTC)

Do you think that the fraction of readers who read "officially concluded" as 'proved' or 'determined as a fact' is negligible? Humanengr (talk) 10:49, 24 April 2017 (UTC)

It is quite difficult to write a nontrivial sentence that nobody will misunderstand. So I don't see that you have point. Zero^talk

What advantage is gained by rephrasing the 1ary source? If both concluded 'with high confidence' and 'officially concluded' leave "a small possibility of being wrong", why rephrase except to distort? Humanengr (talk) 11:53, 24 April 2017 (UTC)

We are the readers of the secondary source; the author of the secondary source is the author. The author has broad discretion about how to write his or her work. If a reader invents a writing rule that the author has no knowledge of, and no obligation to pay any attention to, and the reader then dismisses the secondary source as a distortion, other Wikipedia editors are likely to believe that the writing rule was not invented in good faith, but rather, because the editor arguing for dismissal of the source is grasping at straws to exclude a source that disagrees with the dismissing editor's own views. Jc3s5h (talk) 12:49, 24 April 2017 (UTC)

Two Paper Rule

Might be of interest: https://noahpinionblog.blogspot.com/2017/05/vast-literatures-as-mud-moats.html Nemo 19:17, 23 May 2017 (UTC)

Proposal at WP:PAYWALL (WP:V)

Discussion at: Wikipedia_talk:Verifiability#Encouraging_accessibility

Current

Some reliable sources may not be easily accessible. For example, an online source may require payment, and a print-only source may be available only in university libraries. Do not reject reliable sources just because they are difficult or costly to access. If you have trouble accessing a source, others may be able to do so on your behalf (see WikiProject Resource Exchange).

Suggested (new)

Some reliable sources may not be easily accessible. For example, an online source may require payment, and a print-only source may be available only in university libraries. Do not reject reliable sources just because they are difficult or costly to access. If you have trouble accessing a source, others may be able to do so on your behalf (see WikiProject Resource Exchange).

That said, all else being equal, a source freely available to read online is preferable because more readers will be able to verify its claims. If two sources are equally suitable to verify a claim, accessibility is a reason to prefer one over the other.

Thoughts welcome. Ocaasi ^{t | c} 11:01, 25 May 2017 (UTC)

Military history FA/GA discussion

MilHist project is hosting a discussion on GA / FA articles that have been tagged as possibly having unreliable sources. The discussion can be found here:

• MILHIST: GA / FA articles.

Interested editors are invited to participate. K.e.coffman (talk) 21:46, 10 June 2017 (UTC)

Recent change to Template:Refimprove

We need some opinions on the following matter: Template talk:Refimprove#Or better. A permalink for it is here. Flyer22 Reborn (talk) 01:46, 17 June 2017 (UTC)

Breaking news

The Breaking news section lacks a good definition of "breaking news" and contains a problematic absolutist statement: "All breaking news stories, without exception, are primary sources, and must be treated with caution per WP:PSTS."

I would change that sentence to: "Breaking news stories are sometimes primary sources and should be treated with caution per WP:PSTS."

And I suggest adding this sentence to the beginning of the section:

Breaking news is unedited or real-time reporting about an event as it is happening. Some examples are live video reports, online time-stamped short items that are continually updated, or social media posts. Completed news stories, while they may be updated or corrected later, are not automatically breaking news that should not be used in an article.

- Mnnlaxer | talk | stalk 21:17, 30 June 2017 (UTC)

Um... could you give us an example of a breaking news story that isn't primary? Blueboar (talk) 22:42, 30 June 2017 (UTC)

Not for a breaking news story as I defined it. I would tone down the absolute statement mainly for those that take too wide a definition. But the absolute can stay, the definition is more important. - Mnnlaxer | talk | stalk 01:32, 3 July 2017 (UTC)

So can you give an example where you think someone took to wide a view? (In the abstract, I would prefer to take too wide a view than too narrow... but my attitude could easily change if I could see some concrete examples.) Blueboar (talk) 09:20, 3 July 2017 (UTC)

Diff [1] at Khan Shaykhun chemical attack. Quotes: "it seems that a significant portion of this Wikipedia article's sources come from reputable news sites reporting on what was then breaking news so it is important to remember Wikipedia's policy on breaking news, which can be found here: Wikipedia:Identifying reliable sources#Breaking news" and "Claims sourced to initial news reports should be replaced with better-researched ones as soon as possible, especially where incorrect information was imprudently added. All breaking news stories, without exception, are primary sources, and must be treated with caution per]] WP:PSTS." There are many articles on April 4 and 5. But none that I can see that are primary sources. - Mnnlaxer | talk | stalk 15:03, 6 July 2017 (UTC)

How to characterize biased sources

Are there guidelines for when/how to characterize known partisan sources on politically-related articles? I have seen run into this circumstance several times and am usually met with hostility and lengthy unproductive debates when trying to add information and put sources into context. Here's an example that stands out to me, from James Comey: "His decisions have been regarded by a number of analysts, including Nate Silver of FiveThirtyEight, to have likely cost Clinton the election.[7][8][9]" The "number of analysts" is actually three avowed liberal blogs previously loyal to Hillary Clinton: 538 (mentioned), Vox (which cites 538), and Vanity Fair. As has been established, it's acceptable to use biased sources, but not to sway the tone of the article with these biased sources. So using this case as an example, what would be the preferred re-wording of this sentence? SInce I see no reason to single out Nate Silver, I removed his specific reference. Here a few options I came up with:

• a) "...regarded by several progressive/liberal blogs to have likely cost Clinton the election."
• b) "...regarded by several progressive/liberal blogs to have likely cost Clinton the election. Others are skeptical of this notion and dispute that the letter had any discernible effect." [NYT, Townhall, hotair]
• c) "His handling of the discovery on Wiener's laptop was met with bipartisan criticism.[a few liberal sources and a few conservative sources]"

Without getting too deep into the weeds of what the best option is here (that discussion can be saved for that article's talk page), what would be the best M.O. for adding a frame of reference to sources to ensure that the reader isn't misled into believing that the source is independent and disinterested? Can we possibly create an article with a list of common sources and acceptable descriptors for political articles? I realize that such an article would create debate within itself, but this seems like a common enough problem that it would be worth the trouble. Hidden Tempo (talk) 21:49, 3 July 2017 (UTC)

For additional input on this specific case, the place to go is WP:NPOVN. Appropriate attribution to handle potential bias in sources is a difficult issue, and is liable to prompt disputes, as you've observed. I don't believe it's possible to formulate a list of generally applicable attribution phrases. What we have here are cautiously stated general principles, whose application should be discussed on a case-by-case basis. For example, in a specific discussion I would carefully consider whether "Marxist" is not inappropriately treated as a form of economist's "bias" in a different class from Keynesian, neoclassical, etc, so "should consider" sounds right. Eperoton (talk) 22:33, 3 July 2017 (UTC)

Citing claims in international disputes

When Nation A and A's media accuse Nation B of doing x but those accusations have not been proven in an international court, should WP, in citing A's media, report those accusations as 'fact' or as 'allegations'? E.g., should WP say "B did x" or "A alleges that B did x"? Humanengr (talk) 03:46, 5 July 2017 (UTC)

Nations are generally not taken to court. To illustrate the US was not (legally) at war in Vietnam, but we do not call it the "alleged Vietnam war". So a lot is based upon the context.Slatersteven (talk) 15:47, 5 July 2017 (UTC)

In that example, neither side disputed there was a war. Per WP "In Vietnamese, the war is generally known as Kháng chiến chống Mỹ (Resistance War Against America)." Assume Nation B disputes the allegation. Humanengr (talk) 17:23, 5 July 2017 (UTC)

You mean as in the case of say the "alleged gulf of tonkin incident", or maybe the "alleged Use of chemical weapons in the Syrian Civil War"? Again your question is too vague, there a many instances where there are enough RS saying X is true for us to accept it.Slatersteven (talk) 17:31, 5 July 2017 (UTC)

Let's take your 1st example. Say WP had existed at that point with an article titled ‘Viet Nam attack on U.S. destroyer in international waters”. Should the article rely on media from the US (or allies) or on media from Viet Nam (or allies)? Should the article title have been prefaced with the word ‘Alleged’? Humanengr (talk) 18:45, 5 July 2017 (UTC)

And yet, you are not here for that topic, you are here for Russian interference in the 2016 United States elections — Preceding unsigned comment added by 144.15.255.227 (talk) 18:54, 5 July 2017 (UTC)

The issue is broader, it affects all international disputes. Do you have anything productive to add? Humanengr (talk) 19:01, 5 July 2017 (UTC)

Wikipedia uses the best available scholarly consensus, reporting also where significant bodies of expertise hold differing views. When the US government agencies publicly and unanimously say "Russia did it" and it receives no opposition from any quarter of the academic or other serious political field, Wikipedia presents it as the world sees it - as fact. When Russia claims that they are not involved and all experts laugh, Wikipedia presents it as a laughable claim.

Also, do you care to identify yourself or are you simply here on a fly-by? Humanengr (talk) 19:47, 5 July 2017 (UTC)

We rely on what RS say, and RS tend to be independent of their subjects. Thus a State controlled media is considered less reliable then an independent one. So (yet again) it is not a black and white issue, it is all based upon context (hence my choices).Slatersteven (talk) 10:25, 6 July 2017 (UTC)

What is it that you think editors are here to do? Evaluate evidence? No offense, but what you seem to be suggesting reads like WP:SYNTH. I could be reading into it a bit too much myself, but case in point, you see what I mean, I hope. DN (talk) 05:48, 6 July 2017 (UTC)

Just to provide another example (of which I'm currently indifferent on either way) is 2017 cyberattacks on Ukraine. Ukraine officials blame Russia, but while independent security experts do this some state-level agency may have been behind it, have not named Russia (though clearly would be at the top of the list). There we have not factually said Russia did it, but assigned whom has made that assertion. --MASEM (t) 12:29, 6 July 2017 (UTC)

I agree, as long as we say "X has said this is the case" we are obeying policy.Slatersteven (talk) 13:37, 6 July 2017 (UTC)

@Slatersteven, re 'Nations are generally not taken to court': agree, and until they are taken to court, allegations remain allegations, whatever the media says. Humanengr (talk) 12:53, 7 July 2017 (UTC)

It is an interesting situation, but generally the concept of "innocent until proven guilty" is not applicable where there is not court (other then of public opinion or the media). That is why (for example) BLP's have stricter rules then BLD's, it is not actually as illegal to accuse a dead person of a crime. And again it is hard to see what you are arguing for, each case is separate and we may have situations were the evidence is overwhelming (such as Syrian (by both sides) use of chemical weapons) but where not case has been heard in a court of law. Neutrality does not mean we have to give both sides of an argument equal consideration, rather we give them consideration based upon their reliability, neutrality and their dissemination. As I said this question is too broad to work in most situations, and I can see causing even more trouble then the current situation (can you image how many historical "facts" have never been tried in a court of law).Slatersteven (talk) 13:04, 7 July 2017 (UTC)

Court of Russian, Syrian, Iranian, Chinese, … public opinion and media? R, S, I, C are 'living' (not 'dead' yet despite U.S. best efforts); should not stricter rules apply to 'BLN'? Corporate media is more reliable than state? Humanengr (talk) 13:26, 7 July 2017 (UTC)

"Corporate" media are less prone to being censored or being government mouth prices (that is why they have to be "independent" and not "privately owned" in policy). And no BLN's should not adhere to a stricter set of rules, BLP rules are there because of the threat of legal actions, not as a kindness or out of a sense of fair play. As to your soapboxing about the US trying to kill people, my point was that we have less strict rules for Biographies of dead people because there is less of a legal minefield.Slatersteven (talk) 13:31, 7 July 2017 (UTC)

It might be good to give a concrete example in actual article space to know what the specific complaint is. If it is the Russian interference article, I don't see anything immediately that looks like accusations stated as fact - the lede is very strong to emphasize this is what the US intelligence organizations have determined, but does not say "Russia did it.". (In fact the only thing immediately that looks off is the fact that the Russian response is delegated to a subsection, where it should be on parity with a top level section as a major player, but that's not an RS issue). --MASEM (t) 13:37, 7 July 2017 (UTC)

thx Masem. Taking that article as exemplar (but the issue is broader), yes the lede is (now) reasonably presented. But the article title and heading of the first section are inconsistent with the lede and present as 'fact' without a prepended 'Alleged' b/c of reliance on U.S. media as RS. Humanengr (talk) 14:02, 7 July 2017 (UTC)

A question on youtube videos

I don't really understand why a YouTube video can't be reliable source. What would the difference be if someone made a video about something, then proceeded to write down all the things said in the video and publish them as an article? Its the exact same content made by the original creator.

2601:18D:680:2B2C:ADA5:83C5:9437:95A5 (talk) 00:32, 11 July 2017 (UTC)