Frequently Asked Questions (FAQ) | |
---|---|
|
WikiProject Spoken Wikipedia | |||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
|
Archives |
---|
1, 2, 3, 4, 5, 6, 7, 8, 9, 10
11, 12, 13, 14, 15, 16, 17, 18, 19, 20 21, 22, 23, 24, 25, 26, 27, 28, 29, 30 31, 32, 33, 34, 35, 36, 37, 38, 39, 40 41, 42, 43, 44, 45, 46, 47, 48, 49, 50 51, 52, 53, 54, 55, 56, 57, 58, 59, 60 61
|
Threads older than 60 days days may be archived by MiszaBot II. |
Contents
Primary, secondary and tertiary sources
"Wikipedia articles should be based on reliable, published secondary sources and, to a lesser extent, on tertiary sources and primary sources. Secondary or tertiary sources are needed to establish the topic's notability and to avoid novel interpretations of primary sources."
- It would appear that under this policy a piece of scientific research, published in a peer-reviewed journal by an expert in the field, is to be less relied upon than secondary sources? Someone who has produced the latest, break-through discovery is overridden by someone else's musing in a non-peer reviewed book that might follow up several years later? Wikipedia is behind the times in many of its science-related topics; we have non-science background editors overriding the latest research due to a call to some populist ideas that were published a decade earlier in a book (some bordering on novels). I do not seek a response to this statement as it has been raised before - and very poorly addressed from what I have read in the archives - however Wikipedia editors that have put together this policy need to think about its long-term impacts on the encyclopedia. (NB: I am not a scientist and am not pushing any particular point of view, but I have noticed the impact of this policy on a number of Talk pages of the science-related Wikipedia subjects - those subjects remain about a decade behind what we know today. Users expect an encyclopedia to be up to date else they begin to rely on other, more current sources for their information (e.g. press releases), with all of its inherent issues.) Regards, William Harris • talk • 20:18, 26 December 2015 (UTC)
-
- I am a scientist, and I do support the statement as it is. An encyclopedia should present mainstream consensus views. A breakthrough discovery will, if it is indeed the claimed breakthrough, very rapidly become mainstream. However, in sciences(as in many disciplines) mistakes are sometimes made, and "lucky" finds are often published. Therefore we should be extremely careful with a single paper showing an (experimental) effect.
-
- However, we do not rely only on non-peer reviewed books. We generally also consider the literature review in peer review scientific articles, review and meta-analysis peer reviewed papers as secondart literature. In fact, I think that many of the books you are referring to would be tertiary sources, rather than secondary sources - so not a preferred source either.
- "Under this policy a piece of scientific research, published in a peer-reviewed journal by an expert in the field, is to be less relied upon than secondary sources" is precisely what it means, and is intentional. It's a perennial "Wikipedia adjustment" issue for academics and professional researchers (and even science grad students) when they first encounter this. Especially in the sciences, academic journals are mostly sought by specialists in the field for the hot new research and ideas, with literature reviews seen as sort of dull "catch-up homework", or dismissed as people who are not cutting-edge themselves just summarizing the real work of the eminent. Academic publication, within academe, is tied closely to credit, reputation, and advancement.
For encyclopedic purposes, primary research papers are low-quality sources for facts, because what they are presenting is untested, unverified, unrepeated hypotheses, experiments/fieldwork, data, and conclusions. They've only been "sanity checked" to the extent that a peer review committee thought them interesting enough and their methodology sound enough to be worth presenting to their readers. Depending on the journal, a paper might have been been selected because it is likely to generate controversy and in-the-field "buzz", not because it is felt certain to be correct.
WP relies on secondary (and sometimes tertiary) sources because they reflect real-world review, acceptance, and integration of what the primary material has presented, generally in combination with other research. That is the analysis, evaluation, interpretation, and synthesis that Wikipedians can't do on our own with this material, or rely upon primary sources for, as a matter of "no original research" policy.
Obviously, peer-reviewed secondary sources like literature and systematic reviews are of more value for science that lower-quality secondary sources like newspapers or monographic books by a single author or team, as Arnoutf touches on, above. "Secondary" and "primary" aren't tied to the genre of the publication, but relate to whether it's presenting something new or something arrived at by review of multiple sources. Whether sources (primary or secondary) are of high value or not is matter of author/publisher reputability, and level of editorial review. It may be helpful to review WP:PSTS for WP's internal definitions of primary, secondary and tertiary for purposes of our content policies and sourcing guidelines. These terms can have radically different off-Wikipedia meanings in different fields, and it's unworkable to try to apply one particular field's internal definitions of them to Wikipedia.
Tertiary publications like encyclopedias, including Wikipedia, are always "behind the times" of the bleeding-edge research, because it often turns out to be wrong, and such publications are not in a position to say ourselves which new ideas and conclusions will prove to be correct according to broader scientific consensus. It's just the nature of the publication, its editors, and its purpose.
PS: I have to write pretty much the same explanation of these matters anywhere from 1 to 10 times per month in various places, and I know other editors do, too. We should probably try to improve the relevant pages to make these points (more succinctly than I can off the cuff like this) to forestall these misunderstandings and necessity to keep explaining them in the same terms.) — SMcCandlish ☺ ☏ ¢ ≽ʌⱷ҅ᴥⱷʌ≼ 01:27, 4 March 2016 (UTC)
- ... but there's a lot of scope for editorial judgment in all this, and a lot of argument about secondary sources is really about notability. There's a tension between notability and reliability here. If, for example, a scientist publishes a piece of bleeding-edge research in the scientific literature, and the BBC then report on it, then the fact that the BBC has reported it is strong evidence that the research is important and needs its own page. But the content of that page should be checked carefully against the published research, because we would need our page to reflect the actual research, not the BBC journalist's misunderstanding of that research. In terms of notability the journalist's article is key but in the context of WP:NOR, the scientist's expertise in their own field should trump the journalist's report 100% of the time.—S Marshall T/C 08:47, 4 March 2016 (UTC)
-
-
- Notability: The requirement that a topic have significant coverage in multiple, independent, reliable, secondary sources – notability – is tangential to everything in WP:NOR about secondary sources, other than this is the policy that defines what they are. And the BBC by itself reporting something does not lend it notability (note the "multiple" component that's required – it's the confluence of multiple RS in providing similar coverage that confers notability). Just in case, it may be worth addressing some other common "notability"+"sources" areas of confusion: A newspaper having a larger circulation doesn't automatically make it more notable; that's often a factor of population, target market, advertising, etc. The reputability of a source is a combination of factors, and circulation isn't a major one; how often it's cited by others is a much stronger indication, as is what other RS say explicitly about that writer/publisher/periodical. Many of the world's worst tabloids have enormous circulations and are very notable (mostly in a negative way).
The research: Oh, I agree. But as a matter of policy, all WP:AIES material must come from secondary sources. The obvious solution is you cite the secondary sources for the claims/conclusions, to satisfy policy, and for the general readership to look at, and you cite the original primary research papers for details in the data, for specialist readers to look at. This is how most of our good science article are already constructed, out of necessity. We don't even have to have a guideline saying to do it this way, it just evolves/emerges automatically. I already address the scientists vs. journos matter, and someone else did it before I did, so I'm not certain why that's still your concern; to wit: "Obviously, peer-reviewed secondary sources like literature and systematic reviews are of more value for science that lower-quality secondary sources like newspapers or monographic books by a single author or team, as Arnoutf touches on, above.". I kind of looks like you are saying "whatever a virologist ever says about virology must always be trusted over what a news source says about, even if the scientist's paper is drawing fire because of methodological errors and the irreproducibility of its results, the scientists claims actually came from their personal blog anyway, and the journalism is based entirely on top-flight systematic reviews, and was put out by a publisher with an unusually high reputation for fact-checking with experts"; but surely that can't be what you mean. — SMcCandlish ☺ ☏ ¢ ≽ʌⱷ҅ᴥⱷʌ≼ 09:25, 4 March 2016 (UTC)
- Notability: The requirement that a topic have significant coverage in multiple, independent, reliable, secondary sources – notability – is tangential to everything in WP:NOR about secondary sources, other than this is the policy that defines what they are. And the BBC by itself reporting something does not lend it notability (note the "multiple" component that's required – it's the confluence of multiple RS in providing similar coverage that confers notability). Just in case, it may be worth addressing some other common "notability"+"sources" areas of confusion: A newspaper having a larger circulation doesn't automatically make it more notable; that's often a factor of population, target market, advertising, etc. The reputability of a source is a combination of factors, and circulation isn't a major one; how often it's cited by others is a much stronger indication, as is what other RS say explicitly about that writer/publisher/periodical. Many of the world's worst tabloids have enormous circulations and are very notable (mostly in a negative way).
- @SMcCandlish: the key reason this has to be explained over and over again lies in what you wrote above: "Secondary" and "primary" aren't tied to the genre of the publication, but relate to whether it's presenting something new or something arrived at by review of multiple sources. This means, rightly, that it's not possible to decide whether something is "primary" or "secondary" based on a simplistic rule, such as the medium in which it was published, but rather that editors have to make decisions in relation to the part of the article that the relevant part of the source is being used to support. So cases have to be taken on their merits and discussed as such, and this will continue to be the case. Proper revision of WP:PSTS should result in an increase in the need for discussion and consensus, not a reduction.
- Take one piece from what you wrote above: For encyclopedic purposes, primary research papers are low-quality sources for facts, because what they are presenting is untested, unverified, unrepeated hypotheses, experiments/fieldwork, data, and conclusions. This is actually circular logic, given that you don't equate "research papers" with "primary", since what makes a "research" journal article "primary" is its being used to support reporting in Wikipedia on new and hence unreplicated "hypotheses, experiments/fieldwork, data, and conclusions". To decide that a particular use isn't justified in a Wikipedia article, you have to check carefully both the material it is supporting and the quality of the relevant part of the source. If this can be explained better, great; but don't expect it to diminish the need to keep on repeating the explanation in particular cases. Peter coxhead (talk) 08:48, 4 March 2016 (UTC)
- I wrote unclearly, but thought it was obvious that I was referring to the untested, unverified hypotheses, data, and conclusions in primary research papers (mentioned in same sentence), not any secondary intro material they might contain. I was engaging in some shorthand here, in response to the OP, not drafting new guideline language. I agree with your premise that "Proper revision of WP:PSTS should result in an increase in the need for discussion and consensus, not a reduction." It just shouldn't be this exact same thing over and over again, this confusion of "well in my field, here's how we do it and here's what we think a secondary source is and what value it has" with how WP operates including in the topic of that person's field. It's so, so tiresome to have to keep covering this, month after month, year after year. The very reason we have guidelines is so we don't have to do that. If we do not address it in WP:NOR and WP:RS themselves there will never, ever be a day when this situation improves, because new editors with the same preconceived notions will always continue to arrive. — SMcCandlish ☺ ☏ ¢ ≽ʌⱷ҅ᴥⱷʌ≼ 09:02, 4 March 2016 (UTC)
- I'm sure that fundamentally we don't disagree, and I won't want to needlessly prolong this thread, but some primary "data" is acceptable in some contexts. For example, when a new species is named by a taxonomist, a diagnosis must be attached for the naming to have effect. Facts in that diagnosis are reliable by definition: "the new species is differentiated from all others in the genus by the possession of a retrolateral tibial apophysis" in the original publication supports "the species has a retrolateral tibial apophysis" in a WP article. However, a secondary source is essential to demonstrate the acceptance of the species by taxonomists generally. This is well-explained in the project pages of various Tree of Life descendant projects, but doesn't prevent the need to keep re-explaining it over and over again to new editors.
- Perhaps I've become weary and/or cynical (which I observe to be a common consequence of spending too much time on Wikipedia), but I doubt the wisdom of trying to achieve ever tighter language in relation to complex issues like this one (and I think it's essentially the same issue as in the recent MEDRS discussion). When an issue is subtle, complex and subjective and needs careful explanation, the resulting text will be greeted with "tl;dr" (just think how often the length of your posts is regarded as self-evidently bad) or "I don't understand it so it must be wrong". Even those editors that do read such text will misinterpret it to suit their own ends. Whatever is written down in WP:NOR and WP:RS, new editors with preconceived notions will always continue to arrive, and unlike us ( :-) ), many of their notions will be wrong. Peter coxhead (talk) 09:37, 4 March 2016 (UTC)
- I wrote unclearly, but thought it was obvious that I was referring to the untested, unverified hypotheses, data, and conclusions in primary research papers (mentioned in same sentence), not any secondary intro material they might contain. I was engaging in some shorthand here, in response to the OP, not drafting new guideline language. I agree with your premise that "Proper revision of WP:PSTS should result in an increase in the need for discussion and consensus, not a reduction." It just shouldn't be this exact same thing over and over again, this confusion of "well in my field, here's how we do it and here's what we think a secondary source is and what value it has" with how WP operates including in the topic of that person's field. It's so, so tiresome to have to keep covering this, month after month, year after year. The very reason we have guidelines is so we don't have to do that. If we do not address it in WP:NOR and WP:RS themselves there will never, ever be a day when this situation improves, because new editors with the same preconceived notions will always continue to arrive. — SMcCandlish ☺ ☏ ¢ ≽ʌⱷ҅ᴥⱷʌ≼ 09:02, 4 March 2016 (UTC)
- Meanwhile, the advanced science and mathematics parts of Wikipedia are excellent and getting better precisely because of our highly qualified editors who are able to understand the latest research and summarise it for us mortals. If they had to rely on secondary sources for all their material, a large part of it would become useless and in the case of mathematics a lot of important information would disappear completely. Zerotalk 00:37, 5 March 2016 (UTC)
- Also note from WP:V, no less a policy page than this one: "If available, academic and peer-reviewed publications are usually the most reliable sources, such as in history, medicine, and science." This seems to be in direct contradiction to the claim here that "Under this policy a piece of scientific research, published in a peer-reviewed journal by an expert in the field, is to be less relied upon than secondary sources". The rules on this page were written in an attempt to define and exclude Original Research (which means original on the part of the editor). They were never intended to prevent editors from seeking out the most reliable sources, as specified by WP:V and WP:RS. Note that we also need to observe WP:NPOV, which pretty much covers the gap. If a newspaper article contradicts a paper in Nature without explaining why, we can just ignore the newspaper. However if the newspaper article documents a controversy over the Nature article, quotes other scientists as disagreeing, etc, then we should include that material as well. In all cases, the idea is to present readers with an accurate and balanced summary of what all the experts are saying. Zerotalk 02:20, 5 March 2016 (UTC)
- Exactly so, and well put. It's necessary to make balanced judgements, using sources that are most reliable and appropriate for the material being supported by them. No simplistic rule like "journals bad, encyclopedias good" can replace editor judgement and consensus within the over-riding requirements, such as WP:NPOV. Peter coxhead (talk) 09:53, 5 March 2016 (UTC)
-
Sections for discussing types of OR to serve as cautionary examples
Brought this up at Wikipedia_talk:Editors_are_not_mindreaders#Use of short-cut but I figure it's more related to the purpose of this policy... if we observe other editors engaging in what looks like OR by inserting unsourced claims, it would be very useful to be able to have some kind of shortcut to link to explain their particular problem to them.
Like not that it's the broad concept of OR, but a specific kind of OR so that they can better identify their edit with examples which more closely relate.
Like one problem I experience is one editor keeps claiming a fictional yellow ogre "likes to hug". It seems simple enough. If I had to guess at the character's motives I would think that too. He often initiates hugs with others with a big smile, after all. But it's still OR to assume that he likes what he's doing. To make that claim and not have it OR I think we would need a reliable source (like maybe a char description via a distributor/creator) which explains that he likes hugging. Those would be experts, but simple show-watchers would not be experts.
Basically if it's a type of OR where someone makes unsourced assumptions about how a character feels, what they like, I want to call it WP:EMPATHY. Whereas if it's a type of OR here someone makes unsourced assumptions about what a character's opinion is, what they are thinking about, that would be WP:TELEPATHY.
We already have something like this in WP:BALL (aka CRYSTAL/FUTURE) where users are warned against believing they can predict the future. This relates to cognitive distortion#Jumping to conclusions in this case ball/crystal/future relates to the second one, Fortune-telling, while what I am proposing is a section discussing the first problem, which is Mind-reading. Both are things Wikipedians should avoid doing so both should be warned against.
I believe the WP:RUMOUR (aka SPECULATION) tag could be appropriate to represent a larger section, seeing as how warning against rumour-mongering/speculating warngs against more ills than making guesses about the future, but also guesses about the past/present. In reation to the above CD concepts, I think these two tags should represent the broader taboo of JTC (jumping to conclusions) while BCF represets fortune-telling and WP:EMPA/WP:TELE could point to a new section warning against mind-reading thoughts or emotions.
In this case I am not sure if it is "what Wikipedia is not" or "what Wikipedians are not". The use of BALL/CRYSTAL avoids personifying Wikipedia by saying it is not the TOOL of a fortune-teller. (ie it is not saying Wikipedian is not a seer, but rather it is not the tool of seerS). I do not know any tools of an empath/telepath to substitute here though. At least not one easily identifiable. Even if the focus is on what we ought not to do I think we could still fit it into the WWIN page. 184.145.18.50 (talk) 13:56, 9 February 2016 (UTC)
- This isn't a big enough deal that it would need to be covered in NOR policy. Policies are general, and are not meant to drill down in detail to every conceivable example. If what you're talking about is a common case (I believe it is), it would be better addressed at a topical guideline. I would suggest raising the idea at WT:MOSFIC (and more cogently; it does not help your argument that you are self-doubting yourself throughout it and pointing out why it's not as a good an analogy as you thought it was). It's also not going to help that you're trying to propose multiple shortcuts and are splitting hairs subjectively. It's practical to have a simple proposal to add (at perhaps MOS:EMPATHY in MOS:FIC) a warning against the tendency to imagine and impute to fictional characters various feelings, motivations, beliefs, or intents that are not actually explicit in the work. That could literally be the text of the proposal, perhaps with some examples, that range from obvious to subtle. (And "telepathy" doesn't mean that; it means the alleged ability to mentally communicate over a distance). — SMcCandlish ☺ ☏ ¢ ≽ʌⱷ҅ᴥⱷʌ≼ 09:38, 4 March 2016 (UTC)
Removing citations to "apparently predatory publishers"
unintentional forumshopping Jytdog (talk) 17:04, 3 March 2016 (UTC)
- 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.
Please weigh in: Talk:Predatory open access publishing#Removing citations to "apparently predatory publishers". fgnievinski (talk) 16:16, 3 March 2016 (UTC)
- Fgnievinski I've followed behind you at WT:MEDRS and at JzG's Talk page and at that article Talk page. That section is not an appropriate use of an article Talk page and I closed it. The issue has been under discussion at RSN since Feb 26 here. Jytdog (talk) 17:04, 3 March 2016 (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.
Implied conclusions
If one source claims the first queen of England was Mary I and is on the Mary I page, and another source claims it is Elizabeth II on the Elizabeth II page, would it count as implying a new conclusion to add the Mary I source on the Elizabeth II page with "Mary I preceded Elizabeth II as queen, and therefore Elizabeth II was not the first" because the Mary I source does not specifically say "and therefore Queen Elizabeth II was not the first queen of England"? That would seem like stretching the spirit of this caveat to me, but a pedantic reading would discount using the sources in this way. Yb2 (talk) 20:55, 9 March 2016 (UTC)
- That would indeed be original research. Such combinations are explicitly addressed in the section wp:synth. (PS I guess you mean Elizabeth I not the current queen) Arnoutf (talk) 17:40, 10 March 2016 (UTC)
Especially since there are yet other sources that claim that the first Queen of England was Matilda. Blueboar (talk) 19:06, 10 March 2016 (UTC)
Semi-protected edit request on 11 March 2016
jhk