Wikipedia talk:Reliable sources

NYTimes, NBC and WaPo

Nobody wants to discuss why supposedly reliable sources publish verified news from anonymous sources one week and then retract it the next week? — Preceding unsigned comment added by 2601:844:4000:f910:18c0:6d81:cba5:b07b (talk • contribs)

I do not think they use anonymous sources. Their reporters talk to people they know but who will not let their names be used in print/ The journalists evaluate their credibility before publishing. Rjensen (talk) 03:10, 7 May 2021 (UTC)

This - they will not directly name their source in print but they internally know who they are talking to, and typically identify "according to our source, who wished to remain anonymous..." or the like. As long as the publication is reliable, we presume they are following proper journalist confidentiality here. --Masem (t) 04:07, 7 May 2021 (UTC)

But why did these publication decide just the next week that their sources weren't credible after all? They've published misinformation, and made a big impact with it, and then retracted it within just a few days. — Preceding unsigned comment added by 2601:844:4000:F910:18C0:6D81:CBA5:B07B (talk) 04:43, 7 May 2021 (UTC)

The fact that they retract things is a point in their favor, not a point against them. When we deprecate a source, one of the big things we look for is failure to issue retractions. Hypothetically if it got to a point where a publication were having to issue major retractions every month, then maybe that would stop counting in their favor. But short of that, it's a sign of good journalism. -- Tamzin (she/they) | o toki tawa mi. 05:02, 7 May 2021 (UTC)

A sign of good journalism that they failed to properly vet their sources and had to retract within a week? Doesn't it become a ploy to publish sensationalist fake news to generate traffic and then retract it later? How does this count in their favor? Plus, it must be hard to keep track of all the falsehoods. Here's a story from CNN that still says Officer Sicknick was killed at the Capitol [1]. No retraction in sight.— Preceding unsigned comment added by 2601:844:4000:F910:18C0:6D81:CBA5:B07B (talk)

It's good journalism. You picked a bad example, CNN relied on open-source court papers not anonymous sources. Several days later it reported that an autopsy showed the policeman died of natural causes not the injuries. Rjensen (talk) 22:45, 7 May 2021 (UTC)

So then The Daily Caller or The Blaze, for example, could improve their reliability ratings by retracting claims in their articles after a week? Is that the way it works? — Preceding unsigned comment added by 2601:844:4000:F910:FCEC:F3F6:74FB:45A5 (talk)

Read again from Tamzin: Hypothetically if it got to a point where a publication were having to issue major retractions every month, then maybe that would stop counting in their favor. Please sign your posts.—Bagumba (talk) 02:27, 8 May 2021 (UTC)

See above. The NYT's last major retraction was just last month. 2601:844:4000:F910:FCEC:F3F6:74FB:45A5 (talk) 05:42, 8 May 2021 (UTC)

Every month, not Within the past month. Either you need to start reading comments more carefully, or you're trolling. If you want to convince people that you're participating in good faith here: Let's define "major retraction" as one which garners widespread coverage in not exclusively media-focused sources other than the retracting publication. (That is to say, not a few publications gravedancing over another's mistake, and not media blogs that chronicle everything.) Using that definition, draw up a list of major retractions by NYT, NBC, and WaPo, and report back your findings. And then we can all discuss whether the frequency is excessive. Otherwise, I don't think there's anything left for me to say here. -- Tamzin (she/they) | o toki tawa mi. 06:27, 8 May 2021 (UTC)

I'm not going to spend hours in busy work, but here's a sampling I put together in a few minutes. My point is that the "reliable sources" in Wikipedia's list are subject to the same movement toward sensationalist misinformation as many considered less reliable. That means arguments based on the reliability of these sources are starting from a false premise that they are always reliable. This has implications for using them without question in articles, as well. The NYTimes recently admitted that it's writers frequently present their opinions as fact. Doesn't this suggest the WP policy for "reliable sources" needs another look?

• May 2021: Giuliani story retractions https://www.cnn.com/2021/05/01/media/washington-post-new-york-times-retraction-giuliani/index.html
• April 2021: Russian bounty retractions https://www.foxnews.com/media/media-critics-erupt-russia-bounty-fizzles-nonsense
• March 2021: WaPo retraction of Trump GA call. Don’t see NYT retraction, but they carried this story, too. https://thehill.com/homenews/media/543271-wapost-adds-lengthy-correction-to-story-on-trump-georgia-call
• January 2021: NYT on its own Caliphate mistakes: https://www.nytimes.com/2021/01/12/business/public-radio-group-criticizes-new-york-times-over-caliphate-correction.html
• NYT errors discussion in 2020: https://www.cjr.org/public_editor/nyt-correction-factual-errors-editors-note.php
• NYT rough year in 2019: https://www.washingtonexaminer.com/news/new-york-times-is-having-a-rough-year-cnn-anchors-lament-recent-mistakes
• NYT botched claims in 2017: https://dailycaller.com/2017/08/09/heres-a-list-of-the-5-biggest-ny-times-screw-ups-this-year/

2601:844:4000:F910:FCEC:F3F6:74FB:45A5 (talk) 15:37, 8 May 2021 (UTC)

Why does so much of this page just copy and paste WP:V?

• WP:SOURCEDEF copies WP:SOURCE
• WP:QUESTIONABLE copies WP:NOTRELIABLE
• WP:RSSELF copies WP:SPS
• WP:SELFSOURCE is word for word WP:ABOUTSELF with minor tweaks (like "article" -> "Wikipedia article")

Aren't duplicated sections like this a problem? They are just minor variations of the other, but it becomes arbitrary which section is linked to people, and sometimes one misses out on a point because they happened to read the different version to someone else. It also makes the PAGs look longer than they are. ProcrastinatingReader (talk) 22:29, 9 May 2021 (UTC)

@Newslinger: Can I get your thoughts on this, as one of the resident experts in reliable sources? Can we streamline the wording either here or on WP:V? I think this page intends to be a complete summary on reliable sources from A to Z, which is fine, but then it'd be nice to see WP:V streamlined for duplication. This currently reads and flows much better than that page IMO. Or turn this back into "Identifying reliable sources" and move things like section 1 (definitions) and section 3 (policy on certain types of questionable sources) into WP:V? ProcrastinatingReader (talk) 13:25, 16 May 2021 (UTC)

• Ok... time for some “history of the page”... We have tried “summarizing” as you have suggested (back sometime around 2007 to 2010)... what happened was that the two pages (WP:RS and WP:V) began to drift apart as “instruction creep” took hold. It got to the point where they began to contradict each other. We were constantly having to bring them back into alignment. What we discovered was that the drift was significantly less if we duplicated the text. Since then, the two pages have tended to remain in sync with each other. Blueboar (talk) 13:25, 27 May 2021 (UTC)

Henry Drummond

Regarding my comments on "Henry Drummond": I know that Henry Drummond had at least one child because Henry Drummond is my great-grandfather. His concubine, Karoline Kahn, conceived the child Herman Kahn out of wedlock, therefore there is no documentation. (signed) 73.254.97.23 (talk) 16:41, 11 May 2021 (UTC)Richard Karnes

Disagree: "Although review aggregators (such as Rotten Tomatoes) may be reliable, their audience ratings based on the reviews of their users are not."

Many movies have interesting splits between their Rotten Tomatoes "approved" reviewer aggregated scores, and their audience ratings. These splits are notable in their own right for many movies, and I would argue that because both forms of score aggregation are in-itself aggregations. In the end, there is no discernable difference between them other than elitism. Citing a single audience rating could is problematic as self-published original research, but aggregate is not that, and I would argue, extremely meaningful. I will edit this policy if nobody objects.64.46.20.154 (talk) 11:58, 12 May 2021 (UTC)

See review bombing and similar. Headbomb {t · c · p · b} 12:04, 12 May 2021 (UTC)

The current words are mostly due to some edits in early 2019 e.g. here by RTG. Can you show an article which has been harmed by the sentence's existence? Peter Gulutzan (talk) 13:54, 12 May 2021 (UTC)

Glen Greenwald talks with Tucker Carlson

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.

---

Read the discussion here: [2] K.Q.1997 (talk) 22:54, 13 May 2021 (UTC)

Great journalists like Chris Hedges, Matt Taibbi, and Glenn Greenwald have written that since around probably the 1990s corporatist elitists have controlled most of the mainstream media in the U.S.

Greenwald in particular is great at exposing these people, here is a great article from just today: [3]

There's a lot of literature on the ability to use paranoia and tribal fears in order to manipulate people. You know, when Obama used to be pressured by Marco Rubio and McCain and the hawks in the Democrat party, he used to say that Russia is not a scary power. They have an economy smaller than Italy, they are like a regional power at best and yet in the Democratic party mind, Vladimir Putin is like Darth Vader, Russia is an existential threat.

They've contaminated and infiltrated institutions, they are deliberately stoking the fear constantly among their liberal flock because doing so keeps them frightened, and keeping them frightened means that they are more submissive and more malleable to control. K.Q.1997 (talk) 23:27, 13 May 2021 (UTC)

I propose we close this discussion.

It might also be useful to work on blocking K.Q.1997 per WP:NOTHERE, WP:BATTLE, WP:IDHT, and WP:ARBAP2. --Hipal (talk) 00:23, 14 May 2021 (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.

From BIASED: "Although a source may be biased, it may be reliable in the specific context."

What does this sentence mean? More generally, if I am reading it right, does it imply that biased sources may only be reliable in specific contexts? For context, I feel like the usual interpretation of policy is that bias alone cannot disqualify a source, and was considering WP:BIASED in the context of an RSP discussion for a site with extremely strident, fringe-y views. When I reread this sentence, my brain locked up for a moment because it seems to be written from the assumption that biases disqualify sources by default and that only a "specific context" can salvage them and make them usable. Am I misinterpreting it? Should it be reworded or eliminated, or should it be expanded on to discuss how strong biases may or may not limit the usages of a source and what contexts are appropriate to use them in? --Aquillion (talk) 05:56, 17 May 2021 (UTC)

Good questions. I don't know the original intent here, so I'll just offer some thoughts for consideration.

When the opposing biases found on both sides are mild and close to center, they may exist without ignoring or distorting the facts, and thus both may be truly "reliable" and usable.

As one gets further from center, biases tend to distort things, leaving neither or only one side enough in touch with reality to be considered reliable. That seems to be the explanation, at least during the Trump era, for why any Trump-friendly bias immediately rendered the source as likely unreliable and a pusher of his lies, while opposing sources with strong bias were still fairly accurate and usable. Those sources were still judged by fact checkers to be reliably accurate, even with their biases.

So here's the relevant connection to this thread. If a source is factually accurate in a specific context, then any bias is irrelevant and the source can be used in that context.

Some biases are good because they are still tethered to facts, while others are disqualifying because they let their agenda trump the facts. That's what Trump did and still does. Facts are an inconvenient nuisance to him, so any source favorable to him is suspect, and no context can change those lies to truth. -- Valjean (talk) 06:38, 17 May 2021 (UTC)

@Aquillion, imagine:

• A state-controlled newspaper says that Paul Politician is now the Big Guy at the Department of Administrative Affairs.
• Paul's mother self-publishes her joy at his promotion on social media.

These are situations in which we might have a biased or otherwise unreliable source that could be reliable in the specific context of the claim being made in the article. State-controlled media is pretty reliable on the question of party and government titles; you might not want to trust anything else in the article, but it's probably got those facts right. One's own mother is biased but probably (reasonably) honest.

Since you were looking at a source with strident and fringe-y views, it's probably more relevant to remember that all sources are reliable for something (e.g., a self-published anonymous social media account that published "Dewey Defeats Truman" would be a reliable source for a statement that said "One social media account published the following words: "Dewey Defeats Truman"), and that the general problem with fringe-y views is that they are WP:UNDUE, not that you can't find some nominally reliable source that thoughtlessly repeated them. WhatamIdoing (talk) 03:23, 27 May 2021 (UTC)

My problem is that the rest of WP:BIASED does not contain anything that implies that a source can be rendered unreliable solely due to bias. This sentence implies that that is possible (by indicating, implicitly, that there are some biased sources that are, due to their bias, only reliable in a specific context.) If that's genuinely the case then it should be made more clear and the idea of contextual reliability from bias ought to be elaborated on. --Aquillion (talk) 04:48, 27 May 2021 (UTC)

• I think this sentence is trying to address the fact that reliability isn’t binary - ie Context matters. A source that would be deemed unreliable in ‘’most’’ situations can be quite reliable in a few specific situations (such as when cited to verify a direct quote or close paraphrase).

The problem is that this caveat really has nothing to do with whether the source is biased or not. It is a valid caveat for ANY source that would normally be deemed unreliable - regardless of bias. So it is out of place in a section that focuses on bias. Blueboar (talk) 12:49, 27 May 2021 (UTC)

Think tanks

See Wikipedia:Reliable_sources/Noticeboard#Think_tanks. ProcrastinatingReader (talk) 01:04, 18 May 2021 (UTC)

What if reliable sources contradict each other?

What if a reliable news source published an article that said A=B. But then three peer reviewed papers said A=A.

Would that make the news article questionable or unreliable, even thought the news source itself is considered reliable. I know this sounds weird but, it happens a good amount of times. (Especially around political issues).CycoMa (talk) 15:53, 23 May 2021 (UTC)

A=B does not necessarily contradict A=A. But to rephrase for what I think you're getting at, if media sources say 2+2=5, and all or the overwhelming majority of peer-reviewed academic sources say 2+2=4, we go by the academic sources. This is per WP:Due weight, WP:SOURCETYPES, and WP:FRINGE. The news media articles would then be "undue" or "fringe" in our typical jargon. By political issues, I assume that rather than issues about current events in politics (not usually covered by academic sources and usually covered well by reliable news media), you mean politicized issues, such as climate change, GMOs, etc. In such cases, news media are even less reliable relative to academic sources, because they are far more subject to political bias from their own editorial stance and staff than academic publishing. This applies across the political spectrum. Crossroads ^-talk- 20:44, 23 May 2021 (UTC)

• It's complicated and depends on context. The basic answer is that we look at the relative weight and reliability of the sources. If one source (or group of sources) is clearly better than the others then we just go with them. If they're relatively close, or if both are strong enough and extensively cited that they can't reasonably be ignored, then we cover the difference, but it would still be important to accurately convey the relative weight of the two sides. Another factor is the time when they were published - later sources can sometimes reflect newer information. Peer-reviewed papers are usually better than news sources, all else being equal, but like everything else it gets complicated (three primary studies that haven't been cited many times published in obscure journals vs. front-page articles showing sustained coverage saying the same thing in every major paper of record, say, is not the same as three high-quality secondary studies with massive citation counts vs. one angry opinion piece in a news article.) Another thing to consider that is often worth trying is to read the sources carefully and make sure that they actually contradict, and if so how directly; few things are actually as simple as A=B vs. A=A. Some peer-reviewed papers are hard to interpret, and some news articles imply stuff without saying it. It may also help to look for additional high-quality secondary sources discussing the conflict or contradiction, which might make it more clear if one side was just mistaken, or if the difference reflects updated information, or if one side is WP:FRINGE, or if both sides are treated credibly in academia. --Aquillion (talk) 21:20, 23 May 2021 (UTC)

Okay, I was asking because this is unfortunately very common. Like for example I have seen sources from a very respected news source make claims that went against what peer reviewed and books written by professionals in certain.

The funny thing was the facts from these news sources weren’t technically wrong, it’s the news source and the peer reviewed papers had different interpretations on the matter.

Not to mention the individuals in these news sources appeared to be experts on the matter.(Or at least to some degree.)

But when I did further research on the matter it was very obvious that the views in those news sources are a very small minority.(Too the point they appear to be fringe in a way.)CycoMa (talk) 05:17, 25 May 2021 (UTC)

I wonder if we should address this directly in WP:V and WP:RS. Just a sentence that says "When reliable sources seem to contradict each other, see WP:DUE". Editors often seem to expect us to say "Oh, when there's a contradiction, then that means one of them isn't actually reliable." But that's not how it works: reliable sources can and do contradict each other. Have you ever heard someone say "Two economists, three opinions"? Contradictions are even more likely in the arts and pop culture. What's the music genre? Well, it's whatever the sources say, and if they disagree, then report all of the common claims. Don't look for a rule that makes all except one source get called "unreliable". WhatamIdoing (talk) 03:29, 27 May 2021 (UTC)

WhatamIdoing Thanks for that, I get a basic grasp on that. But some of these contradictions can be extreme.

For example I have seen sources make claims like “experts think this about A” But when I read sources written by experts many of them say something different.

And this isn’t one of those cases where oh Cyco you must have read old outdated sources. Sure that may be the case but in some cases it isn’t.

For example I have seen well respected news sources post articles on why GMOs are dangerous. But, then I would read peer reviewed papers on the topic they say something different. And these peer reviews papers were published years before and years after news articles like that are made.

In weird cases especially when a topic may be controversial, experts from completely different fields of expertise say different things on basically the same topic. A certain view may be controversial to sociologists but, isn’t controversial towards biologists.

I research a lot on topics like gender, sexology, religion, or mythology and trust me you would be surprised how a certain view isn’t controversial among experts in one field but not controversial in the other.CycoMa (talk) 04:00, 27 May 2021 (UTC)

I don't think I'd be surprised, @CycoMa. I think some sources are as prone to engaging in POV pushing as some editors. That doesn't make the sources unreliable; it just makes our work difficult.

Ideally, there'd be a high-quality cross-disciplinary source that says something like "When the psychologists say woman, they mean anyone who self-identifies that way, but when the biologists say woman, they mean any postpubescent human likely to be capable of getting pregnant, and when the lawyers say woman, they mean anyone who is (mis)treated like a woman", or whatever the equivalent is for the relevant subject. In my experience, there are too few such sources, but they are very useful when we can find them. WhatamIdoing (talk) 04:59, 27 May 2021 (UTC)

What if a writer of a source comes out and says they aren’t defending a view

There are situations where I have seen writers in recent years that make articles that claim certain things. But when asked about it on social media the writer admits they aren’t defending a certain view. What do we do about this?CycoMa (talk) 03:12, 31 May 2021 (UTC)

• @CycoMa: - examples? Do we use this information on Wiki? starship.paint (exalt)

Determining if a source is reliable or unreliable

I'm going to make a post without political opinions and commentary because apparently that is not allowed in this section of Wikipedia which I have been made aware of. Please do not close the discussion. I'm only asking questions here in good faith and trying to better understand how Wikipedia operates.

Two questions about reliable sources:

-What are the criteria used for determining if a news source is reliable for usage on Wikipedia?

-What are the credentials Wikipedia editors have that enables them to determine if a source is reliable and/or accurate?

K.Q.1997 (talk) 22:39, 1 June 2021 (UTC)