Wikipedia talk:Identifying reliable sources (science)/Archive 1

Archive 1

Archive 2

Origin of this text in WP:MEDRS

Much of the original version of this page is copied verbatim or closely paraphrased from Wikipedia:Reliable sources (medicine-related articles), specifically the version as it existed on 2010-12-12. That material as well as the text of this guideline are licensed under Wikipedia:CC-BY-SA.

If the method of adaptation is good enough for Pride and Prejudice and Zombies, it is good enough for me. - 2/0 (cont.) 06:40, 12 March 2010 (UTC)

Statement of purpose

Over the next few weeks, I would like to get the obvious problems ironed out before advertising this page a bit more widely and then eventually proposing it as a supplemental guideline to WP:Reliable sources. All are free to contribute collegially. In particular, more examples are needed, especially non-American examples and examples unrelated to physics, the area I know best (disclosure: I am published in Physical Review and am a member of the American Physical Society; the possibility of personal or professional gain through their use as examples here is quite remote). The text of this essay is, and probably should be, overcomplete; reducing repetitiveness and keeping sections focused, though, is probably an area of concern. I would also be unsurprised to learn that I have replicated mistakes long since smoothed over at WP:MEDRS. The parallel between writing medical and scientific articles is close, but imperfect; nevertheless, it would be silly to fail to take advantage of the expertise and perspective of editors there regarding similar matters. - 2/0 (cont.) 06:40, 12 March 2010 (UTC)

Very worthwhile. As a non-expert, may I suggest that " In It can be helpful to create a subarticle devoted to the minority position, leaving a brief summary in the main article" should be accompanied by a mention of the WP:WEIGHT requirement that "such pages should make appropriate reference to the majority viewpoint wherever relevant, and must not reflect an attempt to rewrite content strictly from the perspective of the minority view. Specifically, it should always be clear which parts of the text describe the minority view, and that it is in fact a minority view. The majority view should be explained in sufficient detail that the reader may understand how the minority view differs from it, and controversies regarding parts of the minority view should be clearly identified and explained." I mention it in full because some editors took exception to my attempt to summarise the significance of this provision. This is of course particularly important in articles dealing with a political or social controversy in which science is disputed, and in my understanding the consensus scientific majority view should be shown as such as well as describing minority views favoured by various parties. Don't know if more emphasis is required on showing science correctly even when it is only part of the subject. . . dave souza, talk 22:10, 12 March 2010 (UTC)

Took a stab at it (including that typo - more proof that what we see is not what we think we see) - feel free to tweak or rewrite, as this is meant to be an open sandbox. Much as I may be impressed with my own writing and clarity of expression, there is no chance that this will ever be solid and comprehensive enough to be a guideline if it remains too closely tied to my own parochial viewpoint. - 2/0 (cont.) 19:54, 13 March 2010 (UTC)

Thanks, I think that pretty much covers it, nicely put. . . dave souza, talk 20:08, 13 March 2010 (UTC)

Academic sources, in general

Taking a cue from the development of WP:FRINGE which began as a fringe science guideline and now has developed into a much wider ranging point, I'd like to recommend changing this from Scientific Reliable Sources to "Academic Reliable Sources". This may require more work, but is very worthwhile as many of these ideas relating to sourcing easily cross over into the social sciences and the humanities -- especially disciplines like economics, history, sociology, anthropology, and linguistics.

ScienceApologist (talk) 11:00, 13 March 2010 (UTC)

I can certainly see the utility of such a thing, but I think there may be too much detail lost if we generalize this to something that covers academic papers like Gogol, Pepys, and Andy Warhol: the history of the appearance of references to Wikipedia's controversial "In popular culture" sections in popular culture. Literary criticism and suchlike fields are a very different beast that probably deserves its own sourcing standards; for instance, I think that books have a very different status in more writing-intensive fields without an experimental side. My vastly limited understanding of is that it is a cross between everything is a review and nothing is.

Similarly, economics increasingly uses data-based modeling, but relevance has a much stronger policy component - how many journal articles has Alan Greenspan written? Even in computer science, the advice needs tweaking - that community is much more focused on rapid dissemination and conference presentations.

On the other hand, the advice to prefer scholarly sources and attribute weight to the aspects of a topic according to the sum of the attention of the relevant academic community can be generalized pretty readily. - 2/0 (cont.) 21:27, 13 March 2010 (UTC)

This is long over due.

I saw your post on ScienceApologist's talk page. Creating a variation of WP:RS that addresses the needs of science also interests me. I've been working on a variation at User:OMCV/Sourcing science material. If you do advance this I think it would be good to deal explicitly with all relevant types of sources especially patents. One of the biggest problems is that people just don't understand how science is transmitted. I also think shorter is better than longer, for example I haven't had a chance to read this entire page. I'll stop by again when I get a chance or if you would like me to stay away just mention it.--OMCV (talk) 00:25, 13 March 2010 (UTC)

Took a stab at adding patents. You have a good point about brevity, and trimming for the sake of redundancy or clarity of communication would be welcome. Any other comments and edits are also welcome - this is an open sandbox. - 2/0 (cont.) 21:47, 13 March 2010 (UTC)

Also, please check Wikipedia:Scientific standards. ScienceApologist (talk) 11:02, 14 March 2010 (UTC)

"Grey literature"

I think it might be useful to address the "grey literature" directly - publications that are reviewed, by peers, but not peer-reviewed. Forest Service documents are a nice example - I had a colleague once say that he prefers to publish within the Forest Service, because the review process was much less onerous. Now, obviously, there's a continuum of reliability in the world of the grey literature; at the high end they're as good as a book published by an academic press, while at the low end - I don't know, and that's sort of the problem. Guettarda (talk) 04:22, 13 March 2010 (UTC)

White papers are also of interest. I'd classify them in the same category. ScienceApologist (talk) 11:00, 13 March 2010 (UTC)

WP:SCIRS#White and grey literature - it could use some more examples and the language should probably be tightened, but does that hit the major points? - 2/0 (cont.) 07:26, 26 July 2010 (UTC)

What about reports to Congress/Parliament? These would be unlikely to be mentioned in the peer reviewed literature, but might provide an accessible (both open access and targeted to a non-technical audience) overview of a topic, particularly with regards to any policy considerations. Should we mention them explicitly, or should we just leave it at produced by a widely-cited and respected researcher in the field to avoid stepping on the toes of WP:POLIRS? - 2/0 (cont.) 09:01, 26 July 2010 (UTC)

Wording of sentence

I would like to support this guideline but can't due to this sentence. While I agree with what I think is meant by the sentence, I feel that it promotes original research, weasel words, POV issues in articles and also can't be used for a guideline as the wording "must" would mean this should be submitted as a policy, not a guideline.

I suggest changing this sentence

Although significant-minority views are welcome in Wikipedia, such views must always be described as such and presented in the context of their acceptance by experts in the field.

To this

Although significant-minority views are welcome in Wikipedia, such views should be accorded appropriate weight, and described and presented in the context of their acceptance by experts in the field.

Thoughts?

P.S. For the suspicious amongst us. :) Declaration of interest. None declared. I edit medical articles, governed by MEDRS, so this guideline will not affect my editing. :)--Literaturegeek | T@1k? 18:25, 27 July 2010 (UTC)

It seems like the major issue is changing "must be" to "should be" - is that correct? I don't have a problem with either wording - in fact, "should" is probably preferable to "must" for the reasons that Literaturegeek outlined. MastCell ^Talk 18:40, 27 July 2010 (UTC)

Yes that is the main issue, but also the "always be described as such" is an issue to me as it promotes original research and the "always" is policy type language. Wikipedians don't and can't describe the significance of research as this is original research (not to mention the POV wikilawyering content disputes it would envoke). Only our sources describe the signficance of conclusions and scientific viewpoints, per WP:NOR. Basically I can't support a guideline which appears to contradict one of the 5 pillars, WP:NOR.--Literaturegeek | T@1k? 19:34, 27 July 2010 (UTC)

I made a few tweaks along these lines - see what you think. Exhorting editors to follow the sources for both the raw information and the weighting of it rather than our own intuition or prejudices I think is a major point of this guideline - anywhere you find language that may be interpreted otherwise, please clarify it. - 2/0 (cont.) 13:19, 5 August 2010 (UTC)

Looks good to me. :) Thanks.--Literaturegeek | T@1k? 09:05, 6 August 2010 (UTC)

Scope of the page

I have added a section entitled "Scope" to this page. I believe that this page is meant to apply more to climate change more than it is to apply to Climatic Research Unit email controversy, more to evolution than Kitzmiller v. Dover, and more to superconductor than to Schön scandal. However, the page as it stands does not make it very clear if that is the case. Therefore, how should #Scope be worded to clarify that point? NW (Talk) 14:48, 10 August 2010 (UTC)

Err, I suppose it would help if I read the rest of the page before posting this. I think that a section about the scope of this page would be helpful nonetheless. NW (Talk) 14:49, 10 August 2010 (UTC)

I proposed a scope. Cardamon (talk) 23:27, 18 August 2010 (UTC)

I moved it to the introduction with some text from #When to use, and when not. Feel free to move and improve it back if this works better as a separate section. - 2/0 (cont.) 19:31, 20 August 2010 (UTC)

Wikipedia talk:Verifiability#WP:PAYWALL and freely accessible sources

Editors may wish to take a look at this discussion, which is about the use of paywalled sources. NW (Talk) 02:59, 21 August 2010 (UTC)

This is too oriented towards the Natural or "Hard" Sciences.

Reading this proposed guideline I see that as it stands now it is worded with the hard sciences in mind. Such as Physics, and Chemistry etc. The problem is that this sort of guideline since it just says "science" in the title would be applied to social sciences too.

One problem in the social and human sciences is that the subjects of research often have their own ideas about why they do what they do.

Let me give an example. As it is currently worded a article that was about a tribe of Indians could only use sources written by academic experts, published in peer reviewed journals etc. Very very few American Indians are experts in that sense. Yet they know their own culture better than anyone. The result is outdated, untrue, or even harmful information about a tribe could be put in WP and stated as fact. While sources written by the people under study could not be used, or at best would be treated as suspect and given less weight.

In the social and human sciences the voice of whoever is being studied needs to be given equal weight. (As long as what they write is not patent nonsense.)

Though I am a physicist and have thought that such a sourcing guideline was needed for the natural sciences lately I have spent most of my time editing articles in psychology/sexology. I am aware of the issues of fairness and neutrality that can arise from a policy that favors academically credentialed expertise over confirm able practical knowledge.--Hfarmer (talk) 01:03, 25 August 2010 (UTC)

Nevermind I just noticed the (Natural Sciences) in the title. oops.--Hfarmer (talk) 01:11, 25 August 2010 (UTC)

Looking at the edit history in the time it took me to compose the above the title was changed. :-)--Hfarmer (talk) 01:14, 25 August 2010 (UTC)

Please, let's not revert

I note that there has been an outbreak of reverting in the past several hours. I've requested full protection for this page. I don't know whether or not my request will be approved, but regardless, I urge editors to refrain from warring. Please list in this talk section the passages that are in dispute. I have been saying in this talk for some time that we need to address these issues, or this proposal will go nowhere. If we list the passages that are of concern, here, we can look at ways of finding mutually agreeable language, instead of arguing fruitlessly over language that is unacceptable to some. --Tryptofish (talk) 22:30, 24 August 2010 (UTC)

Note discussion of this page also at Wikipedia talk:No original research#Fresh eyes would be appreciated. --Tryptofish (talk) 23:34, 24 August 2010 (UTC)

And also at WT:V#Fresh eyes would be appreciated. WhatamIdoing (talk) 01:44, 25 August 2010 (UTC)

Seems even typos will get reverted, as also simple use of English grammar <g>. I fear that I shall not partake of this nonsense - it is simply WP:SPOV under sheep's clothing, as any changes to make it in line with actual WP policies and guidelines seem doomed to failure. The wholsesale revert is symptomatic of the issues at the CC arb case now under way. Collect (talk) 23:44, 24 August 2010 (UTC)

Anyone who doesn't want just to take their marbles and go home, some of us still want to work towards consensus. --Tryptofish (talk) 23:48, 24 August 2010 (UTC)

Consensus requires an actual belief of goodwill in making reverts. Where none has been provided, the presumption then falls into the "we know what we want in this, and we will make sure no changes get made" including even typos <g>. I can not take any marbles -- they are all held closely be others who do not want me to touch them (using the analogy given), Collect (talk) 11:05, 25 August 2010 (UTC)

The wholsale reverting of edits by several editors without discussion on the talk page that I could see is yet another reason why it seems like this proposed guideline isn't a good idea. I don't see what could be controversial about saying that the policy trumps this guideline when it comes to a final decision on sources. Cla68 (talk) 00:05, 25 August 2010 (UTC)

Maybe the fact that "policies always trump guidelines" is not actually Wikipedia's policy?

To give you an example, Wolfkeeper has spent the last two weeks telling me that (his carefully selected bits of) this policy always trumps this guideline, and consequently an external link to a comic strip must be kept, because there is no consensus to remove it (or, he conveniently forgets, to include it). WhatamIdoing (talk) 01:38, 25 August 2010 (UTC)

I'm disappointed that there are editors who revert instead of taking some parts of an edit and try to improve it. And I'm disappointed that there are editors who appear disinclined to come to talk. And I also realize that editors come here with past histories. I can't do anything about any of that, of course. But I can promise that I am willing to work in good faith with whoever wants to talk here, to try and find consensus. --Tryptofish (talk) 22:00, 25 August 2010 (UTC)

Interaction with existing policy and guideline

In writing this proposed guideline, one should consider its interaction and compatibility with the policy WP:V and the guideline WP:RS.

For the proposed part regarding the questionable reliability of news reports for science articles, one should consider this excerpt from the section Reliable sources of WP:V,

"Academic and peer-reviewed publications are usually the most reliable sources where available, such as in history, medicine, and science, but they are not the only reliable sources in such areas. Material from reliable non-academic sources may also be used, particularly if it appears in respected mainstream publications."

It seems like this proposed guideline is trying to be more restrictive than this statement, so either this proposed guideline needs to be changed or this excerpt from existing policy needs to be changed or supplemented.

Regarding the section Reliability in specific contexts of the existing guideline WP:RS, there isn't a subsection on the topic of this proposed guideline, so editors here might consider adding one.

Just a couple of suggestions that you can take or leave as you wish. Regards, --Bob K31416 (talk) 17:30, 2 September 2010 (UTC)

A bad precedent to set

I have serious concerns about the propriety carving out a separate sourcing guideline to deal with any one topic. It sets a very bad precedent. This will end up with every topic area defining its own sourcing guideline... which in turn will result in conflicts and disputes (especially in topics that fall within two areas). We could easily justify a separate WP:IRS (history-related articles)... which definitely would be different from the guidelines spelled out here. Yet we have articles like History of science that could be classified under both topic areas. So which guideline would we follow?

The idea here is well intended... but I think it will cause more problems than it solves. Blueboar (talk) 17:49, 24 August 2010 (UTC)

You do know that WP:MEDRS already exists, right? NW (Talk) 18:05, 24 August 2010 (UTC)

OK... so the precedent was already set by MEDRS... that simply means you are following a bad precedent (and proving my point by doing so)... what's next: WP:IRS(politics)? WP:IRS(religion)? WP:IRS(pop culture) This is the way to chaos folks. Blueboar (talk) 02:54, 25 August 2010 (UTC)

My thoughts are the same as Blueboar's, plus a couple more. One is the ambiguous relationship between a specialized guideline and a general policy. Is the specialized one intended to override the general one, vs. set up a condition where a source must meet both of them, vs a condition where a source would have to only meet one of them. ? A second thought is the general theme of making these more restrictive then the general policy. While that might be good for certain cases that someone might have in mind (e.g. a contested technical topic) such would be not good for others (e.g. knocking out a source on an undisputed more basic topic. Sincerely, North8000 (talk) 18:13, 24 August 2010 (UTC)

History of science would not fall under this guideline in my view but rather WP:RS as it is not discussing science per se but history of science, thus it is historical rather than scientific even though it is about science. Policy such as NPOV for example, always trumps guidelines. Guidelines are intended to supplement but not replace policies.--Literaturegeek | T@1k? 18:29, 24 August 2010 (UTC)

This "policy trumps guidelines" meme is not actually true. Editors must use their best judgment to determine which advice to apply in a given situation. There should not be any direct conflicts between policies and guidelines, and any such conflicts should be resolved, not "trumped".

IMO this page provides information and encourages editors to choose higher-quality sources for scientific facts. It provides further information (e.g., that newspapers aren't infallible, that old sources may be out of date) without conflicting with any content policy. It explicitly says that history, social impact, etc. are not scientific information. Surely editors commenting here, if given a choice, would prefer to source basic scientific facts (e.g., the temperature at which water boils, the distance from the Earth to the sun) to a science textbook or a good secondary source in a peer-reviewed journal than to a local newspaper or a random website. WhatamIdoing (talk) 20:20, 24 August 2010 (UTC)

I've been wondering whether merging this into WP:MEDRS might not be a bad idea. Much of the information is necessarily the same. WhatamIdoing (talk) 18:39, 24 August 2010 (UTC)

It seems to me to be premature to bury this. I would agree that it would be a bad precedent to try to set special rules for how societal controversies are dealt with if they happen to involve science. But I think it would be useful to provide guidelines for sourcing scientific facts, and for sourcing the consensus of the scientific community, without spilling it over into taking sides between the POVs of the scientific community, and the POVs of scientists with minority views, science skeptics, and people outside professional science. So far, this draft hasn't gotten that right, but it could, and it's not finished yet. --Tryptofish (talk) 23:55, 24 August 2010 (UTC)

I think Blueboar makes a good point. I could see editors trying to promote a guideline which states that only history books written by professional historians would be acceptable sources. Cla68 (talk) 00:00, 25 August 2010 (UTC)

I can't. Perhaps my imagination isn't as good as yours. WhatamIdoing (talk) 01:39, 25 August 2010 (UTC)

Oh, I can definitely see editors trying to promote such a guideline... Thankfully, I can also see a lot more editors telling them "No way, because that would violate NPOV".

Sound familiar? Blueboar (talk) 02:43, 25 August 2010 (UTC)

Really? You can imagine any editor limiting an entire field of study to a particular form of publication (books) and a particular class of authors (professional historians) -- particularly in a field that is well-known for highly regarded amateurs? WhatamIdoing (talk) 05:44, 25 August 2010 (UTC)

• I think it's worth thinking about precedents. WP:MEDRS includes essentially the same general guidance as that proposed here. It's been a guideline for some time now. And nothing apocalyptic has happened. There hasn't been a wholesale decline in the quality of our medical articles (I think it could be argued that they've continued their slow but steady improvement). There hasn't been a noticeable decline in the (appropriate) use of non-academic sources. In fact, we've attracted a donation of professional reviewer and translator effort for our medical articles. I think MEDRS has been a clear net positive (or, at worst, neutral), so it's really hard for me to take seriously the dark predictions about the apocalypse that will come to pass if this guideline is approved. MastCell ^Talk 04:19, 25 August 2010 (UTC)

I agree that WP:MEDRS has been a good thing overall, but medicine is a little different in that we have a disclaimer and the real world implication of providing false medical information (even though we don't promise anything) is far more dire than someone misunderstanding string theory and there's a robust industry that has an economic incentive for misleading medical information. Is there some particular reason we should hold articles on science to a higher standard than articles on history or religion or football teams? Is there an actual problem that this guideline would fix other than taking sides in the POV-pushing on some high profile science articles? SDY (talk) 06:29, 25 August 2010 (UTC)

Honestly, I understand how something like this could be useful in theory, so I'm not entirely opposed to the idea. however, in practice this kind of thing has some problematical aspects. here's a short list of the problems I think are likely to happen (no, make that inevitable, because you can be darned sure that someone sooner or later will try them):

• Scope Drift. basically this means that people will expand the scope of the guideline as needed to give themselves leverage in their own momentary POV battles. I've seen this happen with MEDRS on ALTMED topics - e.g. someone with a serious attitude against Traditional Chinese Medicine or Acupuncture will pull out MEDRS to try to invalidate every source they believe is favorable to the topic, or to pad in excessive amounts of critical sources. It creates a weird bias: Articles are criticized as not being medical topics by using a guideline only intended for medical topics, with unfortunate results.
• Backdoor Policy Revision. It's one thing for a guideline to clarify an underdeveloped aspect of policy (I think that's the true purpose of guidelines, personally). But it's a fine line between clarifying an underdeveloped aspect of policy and redefining policy. This has already happened with the Fringe guideline (there have been extensive revisions of NPOV which were originally arguments on wp:FRINGE that got transferred into policy by editors who wanted to give them more umph). Some of those changes are useful, some not, but the point is that changes to policy should be centralized in discussion on policy pages, not spread out over guidelines and then slowly leached up into policy on a fait accompli basis.
• Broad confusion over application. As this develops, it will become unclear which of these reliable source guidelines applies. For instance, are articles about cutting edge medical research covered under this proposed guideline or under MEDRS? I could make a case either way, and you can be assured of having editors picking and choosing amongst guidelines to find the one that best suits their particular POV.

Policy and guidelines are supposed to follow the wp:KISS paradigm, and an ever expanding set of distinct sourcing guidelines in not simple. If we want to do this, then I think we should probably absorb MEDRS into this (since there will be a huge overlap anyway) and expand this to cover other scholarly fields as well. --Ludwigs2 06:23, 25 August 2010 (UTC)

On "Scope drift": AltMed is still medicine. Our medicine-related articles shouldn't rely on lousy sources to answer factual questions about these things. (And I'm convinced that you will consider non-technical magazine articles, sales brochures, and the like to be lousy sources on points like "Does Gerson therapy actually keep cancer patients alive?" or "Do more than half of adults have thyroid damage?" or "Does shining a yellow light on the patient cure diabetes mellitus?")

That some editors are anti-AltMed POV pushers, and that most AltMed therapies have been demonstrated to exactly as effective as placebo, is not a valid reason to rely on low-quality sources when high-quality sources exist. WhatamIdoing (talk) 06:37, 25 August 2010 (UTC)

Thank you for reaffirming my point, because you've made the same fundamental error that many others make on AltMed topics. There is a huge grey area of practices that are neither scientifically tested, nor scientifically rejected, nor in any obvious way countermanded by the principles of scientific medicine. This includes things like Traditional Chinese Medicine (which has a couple of thousand years of continuous effective practice), SSRI antidepressants (which apparently don't work all that much better than a placebo, and no one is quite sure why they work at all), splints for fractures (which are simple a traditional medical practice that predates scientific medicine), many kinds of herbal supplements (yes, some herbal medicines clearly have bioactive properties)... You equate all alternative medicine with the absolutely stupidest practices you can think of, and then try to leverage whatever (well-deserved) scientific criticisms you can find on those ultra-stupid practices as though it can be applied to all alternative medicine equally; instant scope drift.

I am not suggesting that low quality sources should be used in preference to high quality sources; I am suggesting that editors are not always as wise as one might like, and we do not want to create a guideline that fosters biases across large ranges of articles. The medical and scientific perspectives are in fact perspectives and should be treated as such. I get a very strong sense that you are using the phrase "High-quality source" as a euphemism for "The Truth", and if so, that's a serious problem that we need to avoid. --Ludwigs2 07:54, 25 August 2010 (UTC)

No, I'm not making this error: Lousy sources are lousy sources, no matter what conclusion they come to. If you've got a well-conducted review that favors TCM (and these certainly do exist), then I'm all for that source. If you've got a lousy "women's magazine" article that hates TCM, I'm against it. Nothing I said can be reasonably interpreted as claiming that high-quality sources are always anti-AltMed sources. It is the type of source that matters, not the conclusions that the source comes to. You can (and should) determine the type of source before you know what the source's conclusions are.

The fact remains, however, that we are far more likely to see lousy pro-AltMed sources than good sources that strongly favor (most kinds of) it. An enormous number of AltMed therapies have been carefully studied. There really isn't much "grey" any longer (except to the extent that everything in medicine is grey, and to the extent that some things do not require randomized studies). A few AltMed treatments (Gerson therapy, which has killed people) are known to be actively dangerous. (These tend to get dropped by the AltMed community over time.) A number "treat" purely hypothetical conditions (i.e., diseases that don't exist, or instances of real diseases that the patient doesn't actually have). The majority are no better than placebo -- and placebo really is one of the very best treatments for some conditions. A few seem to be somewhat helpful in certain situations. (These tend to get adopted by mainstream medicine over time.) These are the basic facts -- but we don't need to know this, or to know which category a given approach falls into, to know whether we're sourcing from the academic literature or from a sales brochure.

Let me put it another way: A reader who wants to know whether TCM treats mild depression deserves an article with the same quality of sources as the reader who wants to know whether SSRIs treat mild depression. Both issues should be addressed with solid review articles, not with manufacturer's websites and magazine anecdotes. AltMed does not deserve second-class status in our sourcing guidelines. WhatamIdoing (talk) 17:01, 25 August 2010 (UTC)

Wikipedia articles do not give medical advice. If a reader wants to know what treatments are available for mild depression, Wikipedia will tell him that SSRIs are used for that purpose (with whatever studies we have on its effectiveness) and that TCM is used for that purpose (assuming that it is notable for that purpose, and with whatever studies we have on its effectiveness). Wikipedia must not go the extra mile and say that SSRIs are better than TCM, except to the extent that reliable research says that. Even if there are no scientific studies whatsoever establishing the safety or effectiveness of TCM as a treatment for depression, if TCM is notable for that purpose we include it without prejudgement or any attempt at minimizing it on the grounds that it's not 'scientific'.

If a treatment is notable for a purpose but occasionally kills people, we note both points. readers are smart enough to figure out for themselves that a treatment that can kill people should be weighed carefully against its potential benefits. Chemotherapy sometimes kills people, certain drug therapies can kill people, aspirin occasionally kills people, yet wikipedia does not go out of its way to say these are bad practices.

It's possible that we may be talking past each other here. I have no problem with using the best quality sources available for a particular topic. However, I do have a problem with sourcing guidelines being used to impose a SPOV over and above NPOV. Science is good, but science is not truth and science is not the absolute dictum on all things in the universe, and no scientist would ever say it is. The problem I'm trying to negotiate is how to keep editors from using a misbegotten understanding of science as a club to beat down practices for purely prejudicial reasons - which happens, frequently. what problem are you trying to negotiate? --Ludwigs2 17:59, 2 September 2010 (UTC)

News as source of criticism?

What is meant by News reports are often inappropriate for reporting scientific results or theories, but are useful when discussing the context of science topics, including the social, economic, and political context, and may provide an important source of criticism. (by SlimVirgin)? Presumably not criticism of the science... RDBrown (talk) 13:17, 25 August 2010 (UTC)

Precedent suggests that's exactly what she means. Short Brigade Harvester Boris (talk) 13:30, 25 August 2010 (UTC)

And she is correct. When a note worthy person has criticized the science behind some theory, we are required to mention it to preserve a NPOV. Newspaper articles can certainly be reliable sources for statements about that criticism (especially for the fact that the criticism exists). We can not use a newspaper article to state that the criticism is correct... but we can absolutely use one for the fact that the criticism exists and what the criticism is. Blueboar (talk) 13:54, 25 August 2010 (UTC)

[1] Short Brigade Harvester Boris (talk) 17:13, 25 August 2010 (UTC)

This language is pretty clearly inappropriate and activist to me. It would be one thing to say that news reports provide an important source of context for scientific thought. To say that journalists are an important source of criticism of scientific thought is highly dubious, at least as far as I think most responsible journalists would see their role. More importantly, it implies that we should mine the lay media for material critical of science. While I think some editors do approach Wikipedia that way, it's clearly not a best practice and we should not encourage it. We should encourage the use of lay media for context - be it positive or negative - rather than encouraging the use of the lay media specifically for criticism. MastCell ^Talk 20:36, 25 August 2010 (UTC)

I'm glad to see this specific point being discussed, as it is exactly the kind of discussion I was hoping would take the place of reverting. Looking carefully at the edits that occurred, the previous language was:

"News reports should generally not be used for reporting actual scientific results or theories, but may be useful when discussing the social context or political impact of science topics." This was reverted here: [2].

Subsequently, this language was added back in:

"News reports are often inappropriate for reporting scientific results or theories, but are useful when discussing the context of science topics, including the social, economic, and political context, and may provide an important source of criticism." [3]

Let me list the differences:

1. "should generally not be used" versus "are often inappropriate"
2. "actual scientific results" versus "scientific results"
3. "but may be useful" versus "but are useful"
4. "the social context or political impact of science topics" versus "the context of science topics, including the social, economic, and political context"
5. "and may provide an important source of criticism", as noted above

My opinion:

1. "are often inappropriate" is better, because it is less prescriptive, less rigid.
2. No big deal either way. Including "actual" may be a little bit helpful, because it draws attention to the fact that it is referring strictly to the science, rather than the context.
3. "but may be useful" is better, for the same reason as #1. (Note how the emphases vary with the POV.)
4. The second one is repetitive, but has the advantage of adding "economic". I suggest this compromise: "the social context or political or economic impact of science topics".
5. Leave it out. It's clearly mining for an anti-science POV. Why wouldn't these sources provide support for science, instead of only criticism? I agree with Blueboar that: "We can not use a newspaper article to state that the criticism is correct... but we can absolutely use one for the fact that the criticism exists and what the criticism is." However, this language goes beyond saying that, and what Blueboar says is adequately contained (I think) in the language about social context and political impact.

--Tryptofish (talk) 22:34, 25 August 2010 (UTC)

My opinions:

1. Agree with Tryptofish.
2. Agree with Tryptofish.
3. Agree with Tryptofish. Also, "are useful" somehow seems to hint that news reports are somehow much better than sources that are neither scientific literature nor media reports (like books). (I doubt that other people will read it this way, but this is how it happened to strike me.)
4. Agree with Tryptofish. I very strongly support the inclusion of "economic" or alternatives (e.g., "business"), and even expanding it ("legal" springs to mind). You don't need to rely on the scientific literature about the gross profit margins in the pharmaceutical (or supplement) industry.
5. Agree with Tryptofish. WhatamIdoing (talk) 23:23, 25 August 2010 (UTC)

Agree with WhatamIdoing! :-) --Tryptofish (talk) 23:31, 25 August 2010 (UTC)

________________

The present form of the sentence is,

"Although news reports are sometimes inappropriate as reliable sources for the technical aspects of scientific results or theories, they may be useful when discussing the social or legal context or political or economic impact of science topics, particularly controversial ones."

I suggest changing a part of it from

"social or legal context or political or economic impact"

to

"non-technical context or impact"

because it's simpler, contains essentially the same information, and is easier to read. Its form is consistent with the earlier part of the sentence which mentions "technical" so that it makes the comparison clearer. (See below.) The three "or"s of the current version are somewhat awkward and the listing of some non-technical subject areas of social, legal, political, and economic doesn't seem useful and may not be sufficiently inclusive (for example, art was left out).

The resulting sentence would be

"Although news reports are sometimes inappropriate as reliable sources for the technical aspects of scientific results or theories, they may be useful when discussing the non-technical context or impact of science topics, particularly controversial ones."

Regards, --Bob K31416 (talk) 00:21, 2 September 2010 (UTC)

I think you make a valid point, even though I was the one who pushed that this should come to talk. To some extent, I don't feel that strongly, because I'm becoming pretty sure that this proposal is going to fail, no matter how we tweak it. My main concern with this sentence has been based on my wish to get editors who oppose the whole idea to come to talk, but I'm guessing they'll just say no-to-change. --Tryptofish (talk) 20:58, 2 September 2010 (UTC)

Thanks. I just started looking at this proposed guideline because I saw it mentioned at WT:No original research. My feeling is that a "reliable source" guideline that goes into details and clarification for a particular subject area may be worthwhile. It may fill a need that may not be practically filled in a general policy due to space limitations. But one has to be careful to be consistent with policy. If something worthwhile is not consistent with policy, one might consider improving policy and then including the item in the proposed guideline. --Bob K31416 (talk) 21:31, 2 September 2010 (UTC)

Interest for this guideline

Has interest in this guideline faded? It would be a shame if all of the hard work that went into this guideline was in vain.--Literaturegeek | T@1k? 23:28, 4 October 2010 (UTC)

I've been wondering the same thing. I have a feeling that this has run up against a brick wall of SPOV. I don't know if anyone out there is interested in developing this into something that is not that. What diminishes my enthusiasm is that anything that goes beyond strictly science (see also the current ArbCom CC case) doesn't deserve to be developed further; on the other hand, it may be valid to provide guidance on sourcing strictly scientific information—but that may be a solution in search of a problem. If there are any good ideas on how to address that, I'll be glad to help. --Tryptofish (talk) 00:44, 5 October 2010 (UTC)

FWIW, I've always found MEDRS to be an excellent guideline to focus sources onto the proper ones (focus on scholarly sources vs news, reviews vs. primary studies, etc.); I can only expect that it would be equally as effective in science topics and a great boon for science coverage in general. Yobol (talk) 01:05, 5 October 2010 (UTC)

I think this guideline is dead in the water because it's politically untenable in the current atmosphere. Most (though not all) opposition to this guideline is based on essentially wikipolitical concerns that will not be placated by good-faith attempts to tweak its wording or scope. In light of that reality, it probably makes sense to find other ways to light one's corner of Wikipedia. That said, I have occasionally been accused of excess cynicism. :P MastCell ^Talk 03:50, 5 October 2010 (UTC)

Oh ye of little faith! Your cynicism is getting the better of you. :) I say we try one last time. I am optomistic because several issues were resolved since when the first RfC was set up. I might file an RfC, if no one is in disagreement? It can do no harm?--Literaturegeek | T@1k? 22:53, 5 October 2010 (UTC)

Oh ye of too much faith! If you mean an RfC to adopt the guideline in its present form, please don't bother. Even I would have to !vote against it. I'm receptive to improving the present draft, but it's not ready for prime time in its present form. --Tryptofish (talk) 23:07, 5 October 2010 (UTC)

Wah, you all depress me and make me sad. :-( Ok, just a little sad, I am ok really. :) Why would you oppose this guideline Trypto?--Literaturegeek | T@1k? 23:28, 5 October 2010 (UTC)

#When to use, and when not, still needs more work. --Tryptofish (talk) 23:48, 5 October 2010 (UTC)

Why don't we just stick a "guideline" tag on it now? Are there substantial objections? (Assume I haven't combed the talk page looking for reasons not to accept it as a guideline). --TS 23:43, 5 October 2010 (UTC)

Ah we can't do that. There needs to be a high level of concensus from the community, per WP:PROPOSAL. Many of the objections were addressed during the last RfC but then some new ones kept cropping up. We would need to run another RfC. I think that running an RfC wouldn't be a bad idea, then if a couple of objections were raised then we could address them and then ask people if they would change their vote from oppose to support.--Literaturegeek | T@1k? 23:50, 5 October 2010 (UTC)

Revival

While the current draft is not perfect, it seems like an excellent proposal. After all, each domain of knowledge has different standards of evidence. For example, what constitutes as a reliable source in science and medicine is much different than the humanities. In the first request for comments, it seems that most of the opposition came from those opposed to all guidelines in general, not this one in particular. However, this guideline may be prove to be very helpful to new editors interesting in contributing to science-related articles. Any guidelines, including this one, that could help support the public understanding of science will receive my approval. razorbelle (talk) 01:00, 1 November 2010 (UTC)

It's not necessary for something to be an official guideline to be useful, and used. Why not tag it as a {{guidance essay}}, and then get back to the real task at hand (which is, after all, to write the encyclopedia, not to write the encyclopedia's advice pages).

If at some point in the future, this page is frequently used and cited favorably in discussions, then we could see about re-opening a discussion of its formal status. WhatamIdoing (talk) 01:29, 1 November 2010 (UTC)

Yes, I agree with that. For editors who are new to the issue, there has been a lot of discussion on-Wiki about how to deal with scientific expertise, how to balance the so-called scientific point of view with the neutral point of view, and some editors are sincerely concerned that a guideline such as this one might be used to undermine the latter. --Tryptofish (talk) 16:04, 1 November 2010 (UTC)

Part of the issue is sector-creep. MEDRS has been used to justify edits to any remotely medically related topic, especially in alternative medicine. There is concern that this guideline would have long tentacles that could reach any article with a hint of science in them--which is to say, a lot of them. Ocaasi (talk) 17:02, 1 November 2010 (UTC)

Promotion to guideline?

While this page isn't perfect, I think it is generally of a high-enough standard to be marked as a guideline. However, I think there should be wider community consultation on that matter first. This is an initial RFC to see if there are any major changes that need to be made. If there aren't, another discussion advertised on WP:CENT could be used to convert the {{essay}} tag to {{guideline}}? NW (Talk) 18:51, 20 July 2010 (UTC)

• RfC Comment. One thing that I noticed on a quick read is that the Books section does not address the existence of single-author monographs that look like very serious scholarship but can actually be close to vanity publications, sometimes done for academic job advancement; these can be relatively unreliable or undue. More broadly, I might suggest seeing how the current ArbCom case on climate change works out, and whether it has any bearing on the content of this page, before taking this to CENT for the broader RfC. --Tryptofish (talk) 21:39, 20 July 2010 (UTC)

I added that we determine the reliability and relevance of a source based on how the experts in the field treat it, and added monographs to the self-published paragraph at WP:SCIRS#Books. - 2/0 (cont.) 07:57, 26 July 2010 (UTC)

Thank you for that. In fact, though, I've seen monographs of this sort that are not, strictly, self/vanity published. There are publishing houses that also publish well-refereed books that publish these things for the revenue stream. But I admit that it is hard to word this in a way that is useful for our purposes. Sometimes a scholarly book by a single author or a single lab group is very much a primary source, with none of the validity of a secondary one. --Tryptofish (talk) 17:24, 26 July 2010 (UTC)

Hmm, I see what you mean. WP:SCIRS#Summarize scientific consensus includes the sentence: If an idea is cited by only a few research groups ("few" varying by field), it should receive very little or no weight in our articles. This gets a little at what I think your point is, but to paraphrase MastCell below we would do well to do more to encourage editors to see the forest more than the trees. Reliable sources do not exist in isolation (hey, that is catchy - maybe someone should add it to the page). - 2/0 (cont.) 18:19, 26 July 2010 (UTC)

Ah, but if a reliable source falls in a forest... --Tryptofish (talk) 18:22, 26 July 2010 (UTC)

• Like. Well, I am a fan of this page, obviously. Tryptofish makes a good point that #Books could better emphasize the role of the major academic publishing houses. I would not want to go too far the other direction by implying that all such books should be blindly followed, but mentioning external editorial control first and then the subtypes of books might work. - 2/0 (cont.) 20:57, 21 July 2010 (UTC)

• Suggestions: I think it's worth carefully reviewing the wording about using primary sources, because (at least with WP:MEDRS) this has been the most heavily wikilawyered aspect of the guideline. Several editors who (to put it kindly) have difficulty with forest-tree discrimination have gotten hung up on the idea that primary sources are never permissible, and that all sourcing has to be review articles. Occasionally this has resulted in sprees where large volumes of appropriate journal citations have been removed "per MEDRS". Obviously, no policy can be proofed against misuse, but ideally we should emphasize the role of primary sources. That is, they can be used - just not misused. MastCell ^Talk 21:15, 21 July 2010 (UTC)

I did some reworking under WP:SCIRS#Definitions, but more is probably needed at WP:SCIRS#Respect secondary sources. - 2/0 (cont.) 08:49, 26 July 2010 (UTC)

• Oppose Guideline creep is not an impressive course for WP. Additionally, this appears to be a form of inserting SPOV as a special class -- when just this week unanimous SPOV that H1N1 was going to be a worldwide destructive pandemic - was found to be wrong. Votes from the WHO etc. do not substitute on WP for fact. Nor does disbelief in what WHO avers as fact then become "fringe." Cirrent guidelines work fine. Collect (talk) 11:41, 22 July 2010 (UTC)
  • To clarify then, do you have a problem with WP:MEDRS? NW (Talk) 12:32, 22 July 2010 (UTC)

Curiously enough, my comments apply to this particular RfC. Trying to ascribe positions to me on any other issue is thereful quite without meaning. Cheers. Collect (talk) 14:17, 22 July 2010 (UTC)

A few things: first, H1N1 did become a pandemic. If you don't get that, then you don't understand what a pandemic is. As to its "destructiveness", the relatively low virulence of H1N1 was recognized more than a year ago by the "SPOV" (e.g. PMID 19524497, Cell June 2009, "The mortality rate associated with infection appears to be low... the spread and impact of this swine influenza virus are similar to those of viruses during an average flu season.") The H1N1 pandemic "only" killed about 18,000 people to date ([4]), so I guess that's not "destructive" enough to impress.

If anything, the popular press was far more alarmist than scientific media about H1N1, so I don't see the relevance of your example to this discussion. More generally, I don't think that pointing out isolated instances where you believe scientists were wrong about something is a coherent reason to oppose this particular guideline. MastCell ^Talk 17:13, 22 July 2010 (UTC)

Average number killed by flu in the US per year is about 25,500 [5] The "pandemic" killed about 25% fewer in the US than the average <g>. So yes - that is not "destructive enough to impress." Now as to WHO ... [www.nowpublic.com/health/full-text-whos-declaration-h1n1-pandemic-swine-flu-virus] from June 11, 2009. So as of June 2009, WHO was still promoting this as the "first global pandemic of the 21st century." I.e. the UN official agency was far more alarmist and asserted the pandemic as a matter of scientific fact. Groups can, indeed, be wrong. [6] JAMA noted "In June 2009, the World Health Organization (WHO) declared the 2009 influenza A(H1N1) pandemic and in October 2009, President Obama declared it a national emergency." Seems the President did not heed your news that the news of a pandemic was exaggerated -- though you assure us that in June it was known that it was no worse than ordinary flu years. <g>. In short -- scientists goofed. Simple fact. And this is "on point" because it is also a declaration of scientific fact by a UN agency. Collect (talk) 17:29, 22 July 2010 (UTC)

I feel a bit silly continuing this discussion, but you're incorrect. H1N1 was a pandemic, full stop, so I don't understand why you're flogging that as some sort of exaggeration. You link a WHO statement from June '09 as an "alarmist" smoking gun, but that statement explicitly reads: "On present evidence, the overwhelming majority of patients experience mild symptoms and make a rapid and full recovery." So it seems that even in the sources you've hand-picked, the virulence of H1N1 was accurately described. I don't see a "goof" here. More to the point, even if the medical community had erred on the side of excess caution in the case of H1N1, I fail to see any relevance to our use of sources in natural-sciences articles. MastCell ^Talk 17:52, 22 July 2010 (UTC)

As I noted, having fewer deaths than average does not actually impress me. It did impress the President who cited the WHO report. Your assertion that by June 2009, scientists knew it was a tempest in a peapot is thus not borne out. And the issue that governmental or multi-governmental scientific organizations can assuredly goof is clearly apt. Clear finally? Collect (talk) 18:17, 22 July 2010 (UTC) see also [7] "the Geneva-based arm of the United Nations relied on advice from experts with ties to drug makers in developing the guidelines it used to encourage countries to stockpile millions of doses of antiviral medication, according to the second report. " That is, "experts" were used to justify pre-selected conclusions for the WHO. Collect (talk) 18:21, 22 July 2010 (UTC)

I think I've said everything I have to say on the subject already. I still fail to see any relevance of these concerns to the actual proposed guideline under discussion. Even if one accepts that criticism of specific WHO actions has some bearing, it would apply to WP:MEDRS rather than here. MastCell ^Talk 22:18, 22 July 2010 (UTC)

Agreed. I've asked Collect to clarify in regards to this guideline. Failing that, this oppose should be ignored. Viriditas (talk) 05:43, 23 July 2010 (UTC)

Try this: Try WP:CREEP for one. Second, that it would discourage properly RS material from being used contrary to WP:NPOV. Third, that WP recognizes no special status for scientific opinions as such (the part about groups being citable for opinions as a special class). Fourth, it is improper to say that an opinion properly posted should be "ignored" as a matter of WP policy. Need more? Collect (talk) 11:38, 23 July 2010 (UTC) As for the snarky asides, they should be redacted, rather than trying to decide to "ignore" my opposition. Thank you all most kindly. Collect (talk) 11:50, 23 July 2010 (UTC)

Further The fact that a statement is published in a refereed journal does not make it relevant is a matter of a great deal of concern -- for example anything published in Physical Review or Physical Review Letters is automaticall considered relevant in the Physics community - yet this "guideline" would relegate it to "does not make it relevant." While this may be true of small medical studies for MEDRS, it is not true in "hard sciences." The Neutrality and No original research policies demand that we present the prevailing scientific consensus, is also contrary to practice in Physics. Cutting edge science is built on the foundation of previous research, and paradigms almost always change only slowly is actually quite false. For example, results that have been reported only in conference proceedings or on a researcher's website are unlikely to be appropriate for inclusion except when reported as such in the author's biography is also quite false. Usually these ideas are proposed by serious researchers who pose a question as part of an endeavor to understand the results more deeply: how can these results be understood in terms of the theory they seem to contradict? ? is meaningless in hard sciences. Statements and information from reputable major scientific bodies may be valuable encyclopedic sources. is simply bosh as far as WP is concerned. And this is only the tip of the iceberg here. Genuine "Scientific bodies" do not vote on facts. I would also note that since mathematics is not a "research" science, that absolutely no math articles on current work could ever be written <g>. Collect (talk) 11:50, 23 July 2010 (UTC)

• Every article I have published in PRL or PRB should be used as a source in every article mentioning mumble, mumble material or technique? Sweet, I will get right on it :). Seriously, though, I am trying to understand what you think that that sentence indicates that is not apparent to me.
• I am not clear on your meaning in the next sentence; do you think that it should be clarified to recognize that, for instance, while many people like SUSY there is currently no consensus as to whether it accurately describes reality? I think that if that article fails to leave the hypothetical technical but naive reader with the impression that SUSY is a serious contender but the data for really testing it does not exist yet, then the article will have failed to inform.
• I am afraid I am not sure based on your input how to improve the next two sentences, other than to suggest that they seem to mean something quite different to me than to you.
• Usually these ideas ... This is intended to convey the fact that the common response by a researcher who finds data that seems to contradict a well-established theory is to tweak the theory, not to propose an entirely new paradigm. Well, after perhaps overzealously checking for experimental error; the trend in the early reported values of the charge of an electron is not a shining moment in experimental physics.
• From the APS. I do agree with you that we need to advise caution when dealing with the pronouncements of organizations who only pretend to the trappings of science, though.

More importantly, Collect, are you just trying to shoot this out of the water, or should we try to find some common ground on wording and presentation? Opining that this should not be made a guideline on grounds of WP:CREEP is a perfectly sound reason to oppose, of course. If you would like, we can start new sections here to thresh out meaning. - 2/0 (cont.) 14:34, 23 July 2010 (UTC)

Collect, is your position that scientists frequently get things wrong, so we should balance them with the opinions of fringe theorists? TFD (talk) 15:20, 23 July 2010 (UTC)

Fancy meeting you here. You seem to think that setting up a straw man argument will get a response. It won;t. Thank you most kindly. Collect (talk) 16:49, 23 July 2010 (UTC)

• Support WP is supposed to be an encyclopedia, and articles should reflect mainstream views. While current policies should be sufficient to ensure articles are properly written, a number of persistent editors try to insert fringe views into them. Therefore clear guidelines are necessary. TFD (talk) 14:30, 23 July 2010 (UTC)
• Support per TFD and 2/0. I also think MCs suggestions are good. Verbal chat 17:48, 23 July 2010 (UTC)
• Support per all of the above. Science is never done (and never absolutely certain) ... so when political bodies and or businesses have to act they rely on the current scientific consensus of the moment which may change tommorrow. Likewise, Wikipedia must report the current state of the science (and uncertainties) along with its historical development and not tomorrow's paradigms or unverified cutting edge stuff. Vsmith (talk) 01:28, 25 July 2010 (UTC)
• I still think it would be worth addressing things like white papers, reports from national and international organisations like the FAO, WHO, USDA, etc. ...and, of course, the IPCC. I think it would be worth discussing white papers, research reports and the like. The key, perhaps, lies in the way that they are treated by high-quality sources - the IPCC assessments are accepted, for the most part, in the peer-reviewed literature. The Discovery Institute's in-house journal, on the other hand, not so much. Guettarda (talk) 03:49, 25 July 2010 (UTC)

I agree that it is a good idea to distinguish between, say, the IPCC or NIST World Trade Center report on the one hand and the Bioinitiative Report or ILADS on the other. Maybe a short section on white and grey literature between #Scientific organizations and #Popular press? Or maybe cover it a merged section with #Scientific organizations? I do not really deal with such things myself, but reception by relevant reliable sources would seem the obvious standard. - 2/0 (cont.) 07:14, 25 July 2010 (UTC)

Yes, gray literature is something that needs to be addressed. Otherwise it is likely to be a point of controversy. Short Brigade Harvester Boris (talk) 00:47, 26 July 2010 (UTC)

More at #"Grey literature". - 2/0 (cont.) 08:56, 26 July 2010 (UTC)

• Oppose. I can't support as currently written, because it is unnecessarily prejudicial to news stories and books written by non-scientists. As I've said before, some of the best books on controversial topics, including science, have been written by investigative journalists. Writing a guideline which might be used to exclude important sources like that is not helpful to what we're trying to do here. Cla68 (talk) 23:00, 25 July 2010 (UTC)

I don't think this is an issue because the example you give is not a scientific matter or describing scientific results. The example you give of lets say an investigative journalist uncovering fraud or perhaps a court case or whatever, would not go in a scientific section but rather a section on "controversy". This is often what happens on medical articles and are more governed by WP:RS than WP:MEDRS.--Literaturegeek | T@1k? 17:16, 27 July 2010 (UTC)

My objection stands as I explained below. Cla68 (talk) 14:34, 10 August 2010 (UTC)

• Support per all of the above. Recently there has been a tendency to ignore the caveats in WP:IRS and simply say "it was published in a book so it's reliable." This guideline would flesh out those caveats in the context of science articles. Short Brigade Harvester Boris (talk) 00:47, 26 July 2010 (UTC)
• Support While much of the popular culture content at Wikipedia is useful and interesting, as an encyclopedia, our core business concerns facts based on evidence, so this guideline is desirable in order to clarify just what is acceptable as a reliable source in the natural sciences. Regarding the discussion above concerning possible scientific errors: I see no evidence of a relevant error being corrected other than by sources which would pass this proposed guideline. Johnuniq (talk) 08:49, 26 July 2010 (UTC)
• Reluctant oppose. It is a very well written guideline and I would like to support it but unfortunately I have a couple of concerns with it as currently written. While I believe that in general secondary sources should be used instead of primary sources I share the concerns raised by MastCell regarding wording on primary sources in that it could be misinterpreted by some to be overly harsh on primary sources. My main concern however, is the below sentence,

Although significant-minority views are welcome in Wikipedia, such views must always be described as such and presented in the context of their acceptance by experts in the field.

The bolded "must always be described as such" concerns me a lot. To use the word must, this can't be a guideline but will need to be submitted as a policy first of all. The wording compells wikipedians to decide "the truth" and not only that, tell the reader of the article what is "the truth", via original research. This sentence isn't even talking about fringe or quackery but significant minority views among scientists. What happens if there are no reliable sources saying that such and such a view is held by a minority or small number of researchers, do we wikipedians perform original research and then decide the truth for our readers? Should wikipedians get in repeated content disputes over labeling researchers and research conclusions "minority view" versus "the mainstream view" if no sources are available? I know on medical articles minority views are governed by "due weight"; if a source says such and such is a minority viewpoint and can be sourced then it can be added to medical articles. I feel this wording has been chosen for a small number of controversial scientific articles and it is going to potentially have an adverse effect on a wide range of the thousands and thousands of scientific articles on wikipedia. If the text I have bolded is deleted I will switch my vote to support. The bolded text is unnecessary anyway with this part of the sentence "and presented in the context of their acceptance by experts in the field" which is quite adequate.--Literaturegeek | T@1k? 18:16, 27 July 2010 (UTC)

Good point - a guideline supplementing a guideline supplementing a policy should make its intentions and place in this hierarchy clear. More at #Wording of sentence. - 2/0 (cont.) 13:08, 5 August 2010 (UTC)

Thank you very much 2over0. Switching my vote.--Literaturegeek | T@1k? 07:38, 6 August 2010 (UTC)

• Preliminary support. My main concern has been resolved and there appears to be no opposition to it. I am changing my vote to support. The reason I am making my vote for support as "preliminary" is because I think we need to establish better consensus among editors on disputed points before acceptance. We are almost there.--Literaturegeek | T@1k? 07:38, 6 August 2010 (UTC)

• Preliminary oppose This needs some more fine-tuning. As it stands it seems slanted to default to keeping out content, which is opposite of the general Wikipedia expectation. This leaves it open to wikilawyering: if the research is recent it's too recent; if it's old it's too old; if it's just one review article "it's just one research group". The sentence "If an idea is cited by only a few research groups ("few" varying by field), it should receive very little or no weight in our articles" is an example of this slant; in my experience (particularly in the medicine area) there's often no more 2 research groups involved in a particular line of research (and often just one), which is not surprising considering that it could be less fun to be just verifying other people's findings. II | (t - c) 03:02, 2 August 2010 (UTC)

I think requiring editors to interpret the general guideline to specific cases is going to be a fault no matter what we do here, but where we draw the line of specificity is certainly open to debate. Particularly with "few" I think you may have identified a point that needs work. I think my reasoning when writing that was that we should let the people actually working in a field dictate what our articles report that experts consider important about a topic (with other matters being outside the scope of this proposed guideline, see #When to use, and when not, below). Perhaps the whole point could be reworded in those terms? Certainly "few" will mean something entirely different in the context of Magnetism vs. Meissner effect vs. Oxypnictide.

Also, I really like your change to describe a secondary source as a source which discusses information originally presented by different authors, as it is both clear and concise. - 2/0 (cont.) 13:01, 5 August 2010 (UTC)

• Strong oppose - This seems to be just another incarnation of SPOV, which I also oppose. ATren (talk) 21:09, 6 August 2010 (UTC)

I disagree, there is little to WP:SPOV other than a brief POV argument on the "truth of science". This is a comprehensive guideline and there are lots of good faithed editors who are trying to come up with a sensible guideline. If you do not raise specific issues and join in the discussion then your oppose will not be very "strong".--Literaturegeek | T@1k? 21:55, 6 August 2010 (UTC)

• Oppose pending a definition of "science-related", which is crucial.--Heyitspeter (talk) 21:20, 6 August 2010 (UTC)

That, about the crucial definition, is a very good point. --Tryptofish (talk) 21:24, 6 August 2010 (UTC)

• Limited Support for now. The only part of it that strikes me as necessary is the rejection of mainstream journalism as a reliable source, since popular opinion on any technical subject is always lacking. I guess the question is if we need a sourcing guideline for "technical subjects" which could include things from Neurosurgery to Aerospace Engineering to Nonlinear Algebra to Kabbalah and Music Theory. Broadly, a technical subject is anything where a nonspecialist is not considered to be a reliable source. The restrictions in this guideline should only be used for discussions of technical aspects of the topic. In other words, it should not be used to ban any discussion of the opinions, especially widely held opinions, of nonspecialists. Articles such as evolution are obviously at stake in this discussion, where dissenting opinions of nonspecialists are critical to a general understanding of the topic. The opinions of nonspecialists should be rejected in any technical discussion, but not allowing citations to support the history of the Scopes Trial (clearly an unscientific venture) would be ridiculous. If the purpose of this guideline is to prohibit any non-technical discussion in articles on technical subjects, then I Strongly Oppose it. Our articles on controversial topics in science come across as preachy and self-righteous, and while I think those who reject evolution are complete morons, their opinion is not uncommon and our basic article on evolution (there are other examples, but that's an easy one) that does not address the social and historical aspects of the topic as well as the technical details is incomplete and inadequate. SDY (talk) 03:09, 25 August 2010 (UTC)

• Oppose: As it stands, I cannot see any practical reason for this page - it doesn't really clarify anything at all as far as I can see - mere wp:CREEP without justification. can someone explain why this thing is necessary, and what it's supposed to do? --Ludwigs2 04:49, 25 August 2010 (UTC)

• Oppose -- Every bit as useless as WP:MEDRS. Everything good in here is already covered by WP:V and WP:RS. -- Jrtayloriv (talk) 21:45, 26 August 2010 (UTC)

• Strong Oppose -- What we have here is the worst "special case" misinterpretations of Wikipedia policies driving out all the good precedent. WP:MEDRS has been a point of contention between excessive and essentially fraudulent medical claims on one side and a pretended unctuous concern for patients that really amounts to not wanting to be upstaged by other countries and modalities of treatment on the other. It isn't a matter of science, and we don't need it to be duplicated and then reduplicated here, with a few more prohibitions on simply covering decent published science every time. And WP:NOTNEWS was written to keep wedding announcements and obituaries out - not to say that we have to wait for someone (who, a journalist?) to review every new well-known result, or that we should cite an AP wire instead of a journal because ooooh, that would be a primary source. We don't need some policy telling us that we can't quote or cite the conclusions of yesterday's Nature paper except by some pass-through. It's just plain wrong. Wnt (talk) 19:41, 28 August 2010 (UTC)

The proposal is to convert this essay into a guideline, not a policy. Boghog (talk) 05:30, 7 April 2012 (UTC)

• Strong support – I think the blind application of WP:MEDRS to scientific articles is excessive and inappropriate. Because Wikipedia medical articles are "widely used source[s] of health information", WP:MEDRS is for a very good reason very strict concerning the need to cite secondary sources. On the other hand, especially when describing recent research results outside the medical field that have been published in high quality peer reviewed journals, I think the careful and selective use of primary sources is appropriate in cases where secondary sources are not yet available. Boghog (talk) 15:36, 6 April 2012 (UTC)

A particularly awful case

An especially bad case is reported here of a pretty obviously absurd claim that lay undetected in the Van Allen radiation belt article for over eight years. A reputable scientist (Alex Dessler) was falsely cited by name, but with no source information. The editor responsible was eventually blocked as a troll, but only six years later. The terrible thing is that a large number of Internet sites (? "quaternary sources" ?) repeated the claim in the meanwhile, many verbatim, so that at a glance one might think it was well-supported. Trying to find a genuine primary or secondary source among the many duplicates would have been a considerable chore. Fortunately it was possible to directly contact Dessler, who responded quickly and disavowed the claim, so it was safe to revert. But it will be interesting to see if that nonsense ever disappears from the web, now that it seems firmly established. This example points up the need for each article to have at some significant fraction of its editors who are genuine experts, as sources may be hard to evaluate otherwise, and errors slow to be corrected. Wwheaton (talk) 21:52, 23 February 2011 (UTC)

• The link was archived to here. The actual edit was one unsourced sentence - it just goes to show, if nobody really gets serious about editing an article, wrong information can last a long time there. The real cautionary story here is just that readers need to always be aware that the purpose of Wikipedia is to get them a reference to look up - if their quest stops at reading what a random person wrote, they'll never be error free; they can't be. I encountered something funny along this line with "glucojasinogen"... The thing to remember with this stuff is that the occasional really wacky claim is only the tip of an iceberg, which we call, well, inaccuracy. There's no sense getting all steamed up about one vandal slipping something past us that is obviously wrong if we're not realizing how often there is stuff which is a little more subtly wrong which we haven't reached the point of being able to do anything about. Wnt (talk) 22:59, 8 October 2012 (UTC)

Organic food

While discussing the allowed sources on the article Organic food (now WP:MEDRS), I was pointed on this page. Unfortunately, I can find clearly that these sources cover agriculture and, of my special interest, Organic food. So I like to know if this is covered or not. And if not, is there another guideline applicable? The Banner _talk 18:51, 8 December 2012 (UTC)

I'm not aware of any guidelines specific to agriculture. To some extent, WP:FRINGE may apply to some (but obviously not all!) of the sourcing issues that might go with organic food. --Tryptofish (talk) 00:18, 11 December 2012 (UTC)

Handling of primary, secondary, tertiary sources

It appears that discussion of this statement as a guideline is dead. I was just pointed to this and read it, and wanted to offer a comment. From my perspective, this essay misses the mark with respect to WP:PSTS. That policy appears in the context of the "no original research"" policy WP:OR. Right? All the bolded do not language in WP:PSTS is to ensure that no OR appears in Wikipedia. Any synthesis or evaluative language must come from secondary or tertiary sources, and those sources are always preferred to primary ones, and it says: "Material based purely on primary sources should be avoided." The policies in WP:PSTS govern sourcing in all articles; the guidelines in WP:MEDRS are within WP:PSTS and add additional requirements to WP:PSTS. However in the "basic advice" section of this essay, where use of primary, secondary, and tertiary sources is discussed, that concern about OR has vanished, and the advice that is presented violates WP:PSTS explicitly and in spirit. For example, editors who see themselves as "experts in the field" appear to be invited to create their own syntheses of material from primary sources. And so on. As long as this remains I believe this essay has no chance of becoming a guideline.Jytdog (talk) 14:00, 26 February 2013 (UTC)

Agree. There's quite a lot wrong with this essay. For example:

• Its definition that a secondary sources have "different authors".
• Any statement beginning "Primary sources should be used when " is simply wrong. PSTS policy says "Wikipedia articles should be based on reliable, published secondary sources". The primary sources are an adjunct and not the foundation of the text.
• The "Respect primary sources" section should be deleted in gross violation of policy. Seminal primary research papers may be included in the article as footnotes for the benefit of the reader, but do not in themselves establish their own WP:WEIGHT and can only be used in limited ways such as straightforward statements about themselves rather than for the facts they claim to have discovered.
• The statement "If mainstream secondary sources in a field do not consider a detail or opinion relevant, it may not be appropriate to cover it at that article; such details and opinions may be desirable at an article on a sub-topic or at a separate article" violate WP:WEIGHT which means that if secondary sources consider it irrelevant then so do we -- don't bury it in some obscure sub-article.
• "Try to avoid citing a source having read only its abstract" should be "Do not cite..." That is simply a dishonest citation.

Colin°^Talk 14:36, 26 February 2013 (UTC)

Science by press release in Neonicotinoid article

See here. I posted there, instead of here, because the work of raising this to guideline or higher is not done yet. Jytdog (talk) 12:27, 1 July 2014 (UTC)

Towards consensus acceptance of the guideline, lets discuss

Ok, I have tried to resolve all issues. Please speak now or for ever hold your silence. :) Raise your concerns, support and objections below.

Issue 1 raised by Tryptofish.

Unresolved

I have made some edits to the guideline to try and resolve everyone's concerns raised during this RfC with the view of gaining a consensus. Tryptofish, raised concerns about monographs being used as sources early in the RfC. Wording was changed to try and resolve his concerns but did not mention monographs, I have changed the book section to specifically address this concern. The previous wording was when chosing a book to consider if it was well received in scientific circles but there are thousands upon thousands of sources for most academic field and very very few sources will it be able to determine how a source is received. Perhaps what was meant was how a viewpoint in the source was received. I have added in content regarding this. Please review the book section.Wikipedia:Identifying_reliable_sources_(science-related_articles)#Books Tryptofish also raised other general concerns which I think I have at least partly addressed with my edits. Please review to see if issues are resolved to a satisfactory level.--Literaturegeek | T@1k? 08:35, 6 August 2010 (UTC)

I think that the edits you made about books/monographs were very good, and addressed my concern very well. Thanks! About my broader concern, discussed by me and MastCell just above, I'd like more time to think about and probably revise further about this. --Tryptofish (talk) 15:23, 6 August 2010 (UTC)

You are very welcome. :)--Literaturegeek | T@1k? 20:48, 6 August 2010 (UTC)

• I'm skeptical of discounting monographs - see diff of the updated language, which attempts to call monographs primary sources. Do we have an example of monographs which are problematic? Monographs are simply scholarly books for which the intended audience is other scientists/professionals rather than students, and Google Books supports this as well as a dictionary. These are high-quality sources which are in general fairly similar to narrative review articles. I also have an issue with calling secondary sources "primary" because "they seem primary". If the author is a step removed from the original presentation of the idea/research, then the source is secondary. And really, what reviews are not from single "research groups"? Many university textbooks and narrative reviews are written by single authors. At RS/N a little while ago I discussed the cost-benefits of secondary versus primary sources. These cost-benefits are dependent upon the extent of the discussion in the secondary source, not whether the secondary source is titled "original research", "review", "monograph", or "textbook". II | (t - c) 19:14, 13 August 2010 (UTC)

I'm going to have to give you an answer that would violate OR if it were in article space instead of here in a discussion amongst editors. Please think of it this way: it's pretty well accepted at Wikipedia that if someone creates a Facebook page or something of that sort, saying how wonderful and notable they are, we at Wikipedia do not accept that as a secondary source that establishes notability. We take note of whether a source is intellectually independent of its subject. Here, on this page, I'm pretty sure we would agree that a similar consideration comes into play with respect to citing a lab's web site as a source.

I'm saying, from a lot of personal experience, that academia is full of examples where scientists who want to burnish their professional credentials write monographs that play up the importance of their own work and disregard the work of their competitors, get them "reviewed" by one of their friends, and a publisher publishes it. Generally, review articles in high quality journals are reviewed more substantively before they are accepted for publication (but there can be gamesmanship there too), but I've personally seen this happen a lot with monograph-style books. And I've seen it happen in areas where the author advocates a minority viewpoint that is contrary to the scientific consensus. Their review gets rejected by a rigorously reviewed journal, so they publish it this way.

Now, please understand, I'm not saying that this is true of all monographs, just some. Deciding how to reliably source the scientific literature for a general, non-professional, audience requires some sophistication in judging whether something that is published has actually been accepted by independent researchers. If you look at my diff to which you link, the conclusion drawn is simply that, when using such sources, editors should make the effort to look at a variety of the literature and make sure the source is representative, rather than taking it on faith that it is. I think that's reasonable. I've looked at your Google Books and dictionary links, and find them completely unconvincing. Your own diff argues that there are places where primary sources are appropriate, and may be more appropriate than secondary ones, and that's a different issue than this one. I'm fully aware as I write this that other editors can in good faith be wondering "should we write this guideline this way because Tryptofish says so?" And I wish I had a better answer to that. Maybe we lack the capacity to make this proposed guideline able to distinguish between reliable and unreliable sources. If so, the proposal may need to be rejected as doing more harm than good. --Tryptofish (talk) 20:57, 13 August 2010 (UTC)

I have no problem with original research on talk pages of any sort - it is an integral part of editorial work. In this case we're doing something original to Wikipedia, so we can use "original research" right in the policy page. But that original research needs to be based on some data. Your point would be much more powerful if you could come up with some illustrative examples or some published discussion of the issue - there are meta journals which discuss this sort of thing as well as the field of sociology of scientific knowledge, although you are probably aware that generally such discussion makes scientists very uncomfortable, and you could maybe try hitting up DGG.

I can come up with plenty of illustrative examples of monographs which look like just basic scholarship. I'll admit that in most of these cases I won't have the intimate knowledge to tease out subtle promotional content. However, I can't see us allowing a blanket "unreliable" removal of monographs or categorizing them as primary sources. The issue of monograph authors not reviewing their own work appropriately is best addressed systematically by acknowledging that when a review/book author discusses their own work, they are not secondary sources - when we say we want a secondary source, we mean someone independent of the original presenter, who can therefore (on average) comment more critically. Such secondary commentary can occur in any type of publication; review articles are on average superior secondary sources, but not always. To jump back to my main point: since monograph is often used (arguably correctly) as a synonym for scientific book, saying that these are primary sources or somewhat unreliable ("Monographs by single authors or single laboratory groups are often best regarded as primary sources, as they may not reflect the views of other experts") is inconsistent with the preceding part of the paragraph and awfully confusing - as I mentioned earlier, how many books are written by more than one laboratory group? II | (t - c) 05:16, 16 August 2010 (UTC)

Thanks. As for data, I'm uncomfortable about giving examples, per WP:BLP—I'd basically be saying so-and-so is a bad scientist. The (current) wording in question is:

"Monographs by single authors or single laboratory groups are often best regarded as primary sources, as they may not reflect the views of other experts. If monographs are used as sources they should therefore be accorded appropriate weight and checked against prevailing viewpoints in the relevant field."

I don't really think that this wording is as bad as "a blanket 'unreliable' removal of monographs or categorizing them as primary sources". But I think that you make a very good point about whether or not the source is discussing their own work. It would be a good idea to improve upon the wording I just quoted, by framing it in terms of sources discussing the author's own work. Let's look at how we might do that. --Tryptofish (talk) 17:46, 16 August 2010 (UTC)

Issue 2 raised by MastCell

Unresolved

MastCell raised concerns over over-draconian language of guideline regarding primary sources and concerns that the guideline could be misused by good faithed editors to delete facts cited to appropriately to primary sources or otherwise have a negative effect on the encyclopedia. I agreed and made some changes. Please review this section, Wikipedia:Identifying_reliable_sources_(science-related_articles)#Definitions and see if this issue has been resolved or if other areas of the article need improved wording etc.--Literaturegeek | T@1k? 08:35, 6 August 2010 (UTC)

I merged everything except the actual definitions out of #Definitions. The relevant material is now at #Respect secondary sources. If I have worsened the presentation or brought back material that is less than ideal, by all means revert or fix it. - 2/0 (cont.) 08:29, 9 August 2010 (UTC)

Issue 3 raised by Collect and Cla68

Unresolved

Collect raised concerns that this guideline was part of a POV battle to insert WP:SPOV. Similarly Cla68 raised concerns about this guideline's wording being used to exclude for example investigative journalism and books by non-scientists. I don't agree with Collect's specific points as I feel that scientific articles should be sourced preferentially to the best quality scientific sources. Newspapers and other non-scientific sources are generally inappropriate for discussing matters of science. However, although I think Cla68 misinterpreted the guideline (in that it is not a ban on social and political and non-scientific controversies), I do share some of the concerns on the points raised by Cla68 that this guideline could be misused to eliminate significant social, legalistics (eg court cases) and political controversies from articles and I improved wording to make it clear that non-scientific aspects of a scientific article can be included in an appropriate article section, eg "controversies", "social and economic" sections etc. What we need to aim to do is to avoid is this guideline being misused to promote the universal truth of science but equally we need to avoid the problem of inappropriate sources being used to promote dubious scientific viewpoints (often referred to as pseudoscience) via inappropriate sourcing when discussing scientific facts. We will never reach a state of perfection through such a guideline but please review this section to see if the changes have helped resolve these issues.Wikipedia:Identifying_reliable_sources_(science-related_articles)#Use_up-to-date_evidence A further point is that I think what is needed is a scientific "manual of style" guideline, for example on medical articles manual of style guideline we have clearly stated that articles should have sections on social and economic aspects of say a drug or disease. With the manual of style for medical article, a POV pusher could not say delete article sections of say a court case against a drug company being fined billions or investigative journalists or politicians raising concerns about fraudulent research for being "not scientific". Equally a POV pusher on medical articles could not use newspaper sources to debunk medically sourced established medical facts.--Literaturegeek | T@1k? 08:35, 6 August 2010 (UTC)

To a large degree, the concern I raised about when to use scientific sources and when not to use them is intended to address these concerns raised by Collect and Cla. If we don't address these very thoughtfully, I can just about guarantee that these issues will arise even more prominently when the larger editing community scrutinizes this proposal. --Tryptofish (talk) 15:23, 6 August 2010 (UTC)

True, but equally we have to be cautious that the guideline does not lead to unintended consequences of the promotion of fringe theories. The only other idea worth considering that I have is having a section in the guideline about article sections covering non-scientific aspects of science, eg political, social, economic, controversies, carefully word it (citing WP:RS, WP:NPOV, WP:UNDUE). Remember any guideline or policy can be abused. I am sure you have seen NPOV or UNDUE being cited during content disputes to eliminate significant minority viewpoints and conversely I imagine you have seen these guidelines being misused to insert and/or give undue weight to fringe theories.--Literaturegeek | T@1k? 20:45, 6 August 2010 (UTC)

I'm not arguing for it to be easy to promote fringe theories, quite the opposite. But I'm arguing for clear language (not easy, or I would already have done it myself!) to address the complaints that will come, both from those who favor fringe theories, and from those editors who in very good faith and for very good reasons want to make sure we don't suppress views from beyond the scientific community. --Tryptofish (talk) 20:57, 6 August 2010 (UTC)

Ok, thanks for clarifying. Yourself, Cla68 and collect or anyone for that matter need to point what language of the guideline remains problematic.--Literaturegeek | T@1k? 21:26, 6 August 2010 (UTC)

My objection will have to stand, and I'll give an example why. As the recent Climategate incident indicated, paleoclimatology, or, at least, much of its research and conclusions, remains a controversial discipline. The use of tree ring histories as ancient temperature proxies, in particular, appears to be far from settled science. Not all of the debate over the credibility and issues over the use of tree rings is discussed in academic or peer-reviewed sources. For example, after MM03 criticized MBH98, which produced the hockey stick graph, Nature invited both sides to submit a follow up rebuttal to each other's papers. The journal ultimately elected not to publish either rebuttal. Although the two sides published their rebuttals on their blogs, RealClimate and Climate Audit, these are self-published sources and therefore problematic. Fortunately for us, there is a book which documents the unpublished debate, and that is the book, The Hockey Stick Illusion. This book is not a scientific textbook, it was written by a science textbook reviewer and accountant who self-taught himself on the paleoclimatic research related to the hockey stick graph. Again, by trying to narrowly define what type of literature we can use in science articles, we are going to cause more problems than we solve. Cla68 (talk) 14:31, 10 August 2010 (UTC)

Thank you for your view. For an area (medicine) which has a similar guideline (WP:MEDRS) if such a controversy existed it would still not go in the medical science sections, but rather it would go into an article section called "controversy" or perhaps the "history" section of an article. I think that (as I outlined in my submission to ArbCom) the problem is that there are opposing editors who are POV pushing of the truth which has led to those articles to degenerate into a battleground. On other articles subject matters, such things are usually resolved fairly quickly, by following, WP:RS, WP:UNDUE and WP:NPOV. I do not believe a guideline can resolve the issues on climate related articles and that ArbCom intervention is necessary as they are behavioural. With that said, I do think that there should be a section in this guideline which covers non-scientific content article sections such as controversy, social, economic and politic content and a manual of style guideline for science articles, similar to WP:MEDMOS would be of value.--Literaturegeek | T@1k? 10:16, 11 August 2010 (UTC)

We should all take a careful look at the recent edits. I'm not sure that we will really have consensus for deleting everything that was deleted. I have an open mind about this, and that's why I've been pushing all along for us to deal with these issues now, so that does not mean that I, personally, object to all or most of the deletions. But I wouldn't want to see us get into revert wars over it. --Tryptofish (talk) 13:55, 24 August 2010 (UTC)

Revert wars are underway it appears. :-O--Literaturegeek | T@1k? 18:33, 24 August 2010 (UTC)

Unfortunately. Please see below. --Tryptofish (talk) 23:45, 24 August 2010 (UTC)

It seems that this objection is based on a view that scientists are often wrong and neutrality requires that their views be balanced with those of sceptics from outside the scientific community. Of course many scientific orthodoxies have been overturned: there is no ether, Piltdown man was a hoax and a comet did hit the earth 65 million years ago. But most fringe theories never gain acceptance. We cannot second guess the best opinion on scientific subjects. People who read articles about scientific topics want to read what they would find in textbooks, not alternative theories. There are of course articles in WP describing fringe theories and of course people can always find websites promoting them if that is what they wish. TFD (talk) 19:57, 24 August 2010 (UTC)

I'm sympathetic to both sides of this, actually. It seems to me that we cannot have Fox News deciding the atomic weight of cesium for us. On the other hand, there are lots of areas of content that are important for Wikipedia to cover, where The National Academy of Sciences may well be expressing only one of several POVs. Part of the problem with the current version of this as-yet-unfinished draft is that is does not really differentiate between sourcing scientific fact, sourcing the consensus within the scientific community, and sourcing issues involving science that are matters of debate within society. --Tryptofish (talk) 23:45, 24 August 2010 (UTC)

There should be room for investigative journalism in this or any other topic. An investigative journalist, such as someone on the staff of 60 Minutes, may not be a scientist, but may have something important to say about an event or issue in science. Cla68 (talk) 00:11, 25 August 2010 (UTC)

Yes, that's absolutely true, but at the same time, they would not necessarily be a source about a scientific fact. I'm trying to point out that there's a difference between scientific facts (and the consensus amongst scientists about them), and broader societal issues about ways in which scientific information fits into society. I feel as though editors, who have histories in these editing controversies, are talking past one another about this. --Tryptofish (talk) 00:23, 25 August 2010 (UTC)

Investigative journalism may point out that what one group of scientists is pushing as a scientific "fact" may not be as certain as those scientists say it is. So, instead of ruling out some sources, it should be left open for editors to decide on the article talk page by cooperation, collaboration, and compromise. Cla68 (talk) 00:28, 25 August 2010 (UTC)

What "one group of scientists is pushing" is not the same thing as the current scientific consensus. WhatamIdoing (talk) 01:32, 25 August 2010 (UTC)

Yes, they are not at all the same thing. It seems to me that Cla's point depends on whether the investigative journalists are reporting that there are reputable scientists questioning what the "group" have been "pushing", which is likely to be news that passes WP:N, or whether they are reporting that a group of conspiracy theorists, for example, are questioning it. Investigative journalists report on both these kinds of stories, and Wikipedia needs to be able to know the difference. --Tryptofish (talk) 21:41, 25 August 2010 (UTC)

P.S.: I didn't mean we should never allow the latter, but, rather, that we need to be able to treat it differently, lest we mislead our readers. --Tryptofish (talk) 21:55, 25 August 2010 (UTC)

Issue 4 raised by II

Unresolved

ImperfectlyInformed raised concerns about wording on "few research groups" pointing out that often only a few research groups conduct research, which is true and I agree that it could have an adverse impact on our articles and could be misused by editors to POV push to keep content out of articles. Please review this article section.Wikipedia:Identifying_reliable_sources_(science-related_articles)#Summarize_scientific_consensus--Literaturegeek | T@1k? 08:36, 6 August 2010 (UTC)

I still have some issues with this section. For example, this sentence: "Scientific consensus can be found in e.g. recent, authoritative review articles, high quality journal articles, or widely used textbooks". I've read a few reviews/articles/textbooks and I'm biased to the medical/toxicological (and in the social sciences, finance and economics) perspective. In these areas, it is not my impression that reviews and books regularly use the word "consensus", nor am I sure that they want their readers to come away with a consensus view. I thought that the point behind WP:RS/AC is that we don't call basic scientific findings "consensus" and we don't generally collect a few articles saying the same thing and say that this means consensus. Most scientific facts don't need consensus. The scientific areas where the word "consensus" starts getting thrown around tend to be the ones which are political and somewhat uncertain. Many scientific facts are either so uncertain or so esoteric that there is no consensus except partial agreement among the five-six scientists working in the area, which is not really something they'd call consensus. Anyway, there's a lot of talk about consensus and I think it's confusing. There's also no impression on how to assess this so-called consensus. II | (t - c) 08:15, 11 August 2010 (UTC)

Thank you for view. I think what the consensus section of the guideline is trying to do is to say is give most weight to the dominant view of scientific literature and less weight to minority viewpoints weighted against their prominance in a given field or area of a field. The only solution that I can think of is that that section is merged with other sections in the guideline and consensus reworded into terms of "dominant" and "minority". This would, pardon the pun, require consensus of other editors here. :)--Literaturegeek | T@1k? 10:40, 11 August 2010 (UTC)

WP:MEDRS has a section on scientific consensus which has been part of the guideline for a long time without causing any problems or controversies among editors.--Literaturegeek | T@1k? 18:32, 24 August 2010 (UTC)

Issue 5 raised by Heyitspeter

Unresolved

Heyitspeter raised concerns that the guideline does not define what "science-related" means. Defining this may help resolve some of the remaining issues raised by other editors. I think that it should be defined as scientific content in science related articles of which (scientific content) will naturally be the bulk or entire content of most if not all scientific articles). Criticisms, controversies, legal, socio-economic, politic issues and other non-scientific content issues, when notable can be added to a non-science related section towards the bottom of the article (we still need a science related manual of style guideline, similar to WP:MEDMOS), following the existing guidelines and policies such as WP:UNDUE, WP:NPOV and for sourcing WP:RS. Wording will need to be carefully constructed to make the less likely to be misused to give undue weight to notable minority viewpoints or conversely we have to be careful that the wording is not easily misused to suppress notable minority content.--Literaturegeek | T@1k? 21:36, 6 August 2010 (UTC)

I think that this is largely the same issue that is being discussed at #When to use, and when not. - 2/0 (cont.) 03:52, 7 August 2010 (UTC)

Yes, that's right, it is. I do think that Peter summed it up in an important way by noting the lack of a definition of "science-related". --Tryptofish (talk) 16:58, 7 August 2010 (UTC)

Issues raised by 75.152.127.40 (talk)

1. Reeplaysez direct evidence (primary research) with opinions about much research (reviews). To go into detail on this point, I want it understood that sometimes wikipedia, because of its sheer size, and because you do not need to know (and sometimes not even obey) the rules to edit wikipedia, we are a source of orijinal thought. For example, condense the conclusions of two papers, so that anecdotal evidence sujests that soy beans cause impotence, and an experiment on rats with soy isoflavones sujests that isoflavones are phytoestrogens. However the connection is made in the first place, we do not conclude that soy beans cause impotence. We figure it out from the juxtaposition, and look for synthesis from somebody who should know. If it takes a while for someone to find the synthesis, then primary sources sit on wikipedia for months, perhaps even for the years that a citation request haz remained outstanding. Policy says you can use primary sources. I believe that even after the synthesis is found, that the primary sources should remain: "Soy beans cause impotence (1,2,3)."
2. Haz a nutshell and a template that are too simple. Delete the nutshell. Delete the template. Do not call for action in a terse manner. We do not need any more automatons on wikipedia who are not capable of didactic reasoning or critical thinking beyond other policy wonks on wikipedia.
3. Duz not go into reasons to include primary research, like sheer sparseness of research, or list-quality articles like hepatoprotection. Guidelines really should hav a lot more exceptions to them than policy, to which there should be more exceptions than pillars. If you spell out article types for which this essay should not apply, then it might become a guideline.
4. Haz a bias in favour of reviews that are restricted to RCTs (meta-analytical), which means a bias against human-weighed summaries. (That's what it adds up to on medrs, anyway). "A systematic review uses a reproducible methodology to select primary studies meeting explicit criteria in order to answer a specific question." Yeah, sure, and sometimes it is very difficult to understand how both the inclusion criteria and exclusion criteria add up to excluding honestly constructed fringe experiments.
5. Haz an explicit bias against primary sources that may be the only sources for a fact, meaning that it iz pointless to make jeneral statements about whether crucifers prevent all forms of cancer, and useful to say that some of them prevent colon cancer: A review might be useless for the scope of summary that it iz trying to make.
6. [Do not juxtapose contrasting findings, unless that's what the author intended.] No. Delete. Let arguments sit that way until someone can find out who got tobacco money. In the best case scenario, a third party will come along (with support) and either word two sentences to exclude a conflict, or resolve a conflict in some medium.

75.152.127.40 (talk) 06:10, 15 July 2014 (UTC)

When to use, and when not

I've been thinking about the idea of proposing this page as a guideline, and trying to put my finger on something that has made me uncomfortable about doing so. I'm coming at this as someone who has been a professional academic scientist in real life, and who is very receptive to making sure Wikipedia uses high quality sourcing, but who, at the same time, can easily see both sides in the current on-wiki controversy about climate change. I think this page needs to do a better job of delineating when scientific sources are, and are not, the proper sources to use on Wikipedia. Otherwise, I'm pretty sure that it will fail to achieve consensus, out of concerns that it will be used to game content disputes.

I've pulled out these three passages from existing science-related pages, to show how sourcing needs vary with the content that is being sourced:

• From Parkinson's disease: "L-DOPA is transformed into dopamine in the dopaminergic neurons by dopa-decarboxylase."
• From Autism: "Parents may first become aware of autistic symptoms in their child around the time of a routine vaccination, and this has given rise to theories that vaccines or their preservatives cause autism. Although these theories lack convincing scientific evidence and are biologically implausible, parental concern about autism has led to lower rates of childhood immunizations and higher likelihood of measles outbreaks."
• From (gasp!) Climatic Research Unit email controversy: "The scientific consensus that "global warming is happening and that it is induced by human activity" was found unchallenged by the emails, and there was "no evidence of any deliberate scientific malpractice in any of the work of the Climatic Research Unit.""

The first (despite a bit too much alliteration) is making an unambiguously factual scientific statement, for which scientific sources are clearly appropriate. The second, summarizing scientific consensus in an area where there is a well-established scientific consensus, and where there are serious medical consequences to what we, at Wikipedia, do or do not say to our readers, is also clearly scientific, but nowhere near as clear-cut as the first, and deals with a subject where there is also a notable fringe view. The third, well... sits at the controversial interface between science and broader public issues.

The first sentence can come from a lot of places. Glial cells might make dopamine, too. I hav the vaccination autism noise from a facebook video that says it wuz based upon a single, retracted story in the lancet. I bolstered it in Breast milk with something from Oman that said suboptimal (less than six months exclusive) breastfeeding correlates with autism. Someone deleted it. I do not know why. The third contains a self-referential error that's actually kinda funny.75.152.127.40 (talk) 07:10, 15 July 2014 (UTC)

It's not going to be easy, but I think this page needs to better differentiate amongst these sourcing needs, if it is to be accepted by the community. --Tryptofish (talk) 22:28, 2 August 2010 (UTC)

To be clear, the scientific consensus on climate change is as strong, if not much stronger, than the scientific consensus that vaccines do not cause autism. The difference is that the media has lately done a somewhat better job of conveying the scientific consensus on vaccines, and has increasingly declined to give "equal time" to vaccine critics. As to your question about the CRU, it would seem to me that if we're discussing the scientific validity of specific findings, then scientific sources are the best place to look (in this case, the text seems to be quoting one of the outside reviews of the CRU emails). The scientific reviews occasioned by the CRU document leak were also covered extensively in the press, so we could probably use that coverage as well. If we're discussing the political impact of the CRU document leak, then there's room for both scientific sources and good-quality press sources. MastCell ^Talk 22:48, 2 August 2010 (UTC)

Gasp again: I didn't start this thread in order to have yet another page to argue about climate change. That's not the point, please. But at least I agree with your points about multiple kinds of sourcing being appropriate to that kind of material, and I want to draw attention to how that contrasts with dopa-decarboxylase, and to note the potential importance of not misleading readers into failing to vaccinate their children, without getting into an argument about the relative strengths of scientific consensuses for climate change versus autism and vaccines. Believe me, if we don't deal with this here, it will come up when the page is proposed more widely, as this first response to my post amply shows. --Tryptofish (talk) 22:57, 2 August 2010 (UTC)

No, I don't want to have an argument about climate change either - it just seemed that of the three examples you mentioned, the CRU was the only one that presented much of a gray area. I guess my initial point was that the weight of scientific opinion on a topic is not consistently conveyed in popular-press coverage.

If you go back a few years, the autism articles used to be a hotbed of low-quality, agenda-driven editing. They've calmed down and improved in quality greatly since then - in part due to the heroic efforts of editors like Eubulides (talk · contribs), but also because the popular press no longer caters to the vaccines-cause-autism lobby the way it did a few years ago. It was a lot harder to write good-quality, neutral, encyclopedic material on autism when the popular press was committed to depicting Andrew Wakefield as a heroic Galilean martyr at the altar of scientific orthodoxy. It's easier to write good-quality, informative material on autism now, but the challenges have moved on to other areas.

I guess I'm mostly agreeing with you - the guideline needs to provide some suggestions on integrating popular-press/sociopolitical coverage and sourcing with expert scientific opinion and scientific sourcing, or it won't be very useful. On the other hand, we can't prescribe a rigid formula for balancing the two sources - the most we can do is come up with some general guidance and appeal to common sense. MastCell ^Talk 17:05, 5 August 2010 (UTC)

Thanks for that, my gasp is retracted! Yes, I think we do largely agree, and I want to point out that, of course, there is nothing particularly special about the three examples I tried to use to illustrate what I was saying. But what I do think is very important here is that we need to do more than appeal to common sense. (What, Wikipedians practicing common sense? Impossible!) You are right that we won't be able to prescribe a rigid formula, but I think we either need to provide some guidance about how to use scientific sources along with other-than-scientific ones, or the community will see the proposal as something that has an ulterior motive. --Tryptofish (talk) 19:56, 5 August 2010 (UTC)

These problems are really quite endless. There are fringe theories that are quite right: Avoiding sunlight causes more problems than getting it: Labourers do not get more cancer than the rest of us. Yeah, I know, I'm writing this at night. 75.152.127.40 (talk) 07:10, 15 July 2014 (UTC)

I invoke WP:TENFOOTPOLE on your third example, but agree that this proposed guideline should not be used to exclude or marginalize well-sourced social or historical aspects of articles (this is intended to be supplemental to WP:IRS and WP:V, not WP:DUE). The second paragraph of the lead section already touches on the demarcation problem, and could probably be expanded to note something along these lines. I am a little worried that nobody will read a bloated guideline (currently 38 kB to 35 for WP:MEDRS); some of the points can probably be made more concisely without sacrificing accuracy); it is possible that delineating when this page's advice is relevant can be done by tweaking the text in #Definitions in addition to or instead of the proposed text below.

I'm from Alberta. I hav facebook friends who work in the industry of climate change. I like David Suzuki. I should've invoked tenfootpole a long time ago. 75.152.127.40 (talk) 07:10, 15 July 2014 (UTC)

To reminisce about the incident that first spurred me that we need something like this page, about a year ago the popular press and much of the science journalism press completely mangled reporting the results of a Science paper on magnetic monopoles (see: Talk:Magnetic monopole#recent monopole 'discovery'). In such a case, a quality encyclopedia has to ignore the chaff, instead reporting the research itself. It is a fine line to walk between giving general advice on how to search for and identify reliable sources for this context without encouraging original research.

Diamagnetism merjer, anyone? 75.152.127.40 (talk) 07:10, 15 July 2014 (UTC)

Proposed text for second paragraph of the lead section:

No source is universally reliable. Each source must be carefully weighed in the context of an article to judge whether it is reliable for the statement being made and is the best such source. This guideline provides advice for finding and identifying sources reliable for statements made about current or historical research in the natural sciences and closely related fields. It does not cover reliability in context of the social sciences, biographical detail, social or political impact or controversy, or related topics that are sometimes appropriately covered in natural science articles.

I hav a whole textbook to read from an expert, and I anticipate that it will piss me off, because he is a specialist, and I do not believe him. 75.152.127.40 (talk) 07:10, 15 July 2014 (UTC)

This may be overly restrictive, but does it look like a step in the right direction? - 2/0 (cont.) 05:44, 9 August 2010 (UTC)

Yes, it does, as do the other edits you have made. I think that what you wrote about not using non-science sources to determine scientific consensus is particularly helpful. (Only ten feet?) --Tryptofish (talk) 18:33, 9 August 2010 (UTC)

More than ten feet; Kilometres: We call them clicks here. 75.152.127.40 (talk) 07:10, 15 July 2014 (UTC)

Beall’s List of Predatory Publishers 2015

FYI: Beall’s List of Predatory Publishers is a list of scientific publishers who will publish any paper, regardless of quality, for a processing fee. In some cases the predatory publishers uses a name that is similar to or identical to the name of a legitimate journal. --Guy Macon (talk) 16:52, 15 August 2015 (UTC)

RfC announce: What does Wikipedia:Identifying reliable sources (medicine) cover?

There is a request for comments at [ Wikipedia talk:Identifying reliable sources (medicine)#What does MEDRS cover? ].

At issue is whether the lead paragraph OF WP:MEDRS should remain...

"Wikipedia's articles are not medical advice, but are a widely used source of health information. For this reason it is vital that any biomedical information is based on reliable, third-party, published secondary sources and that it accurately reflects current knowledge."

...or whether it should be changed to...

"Wikipedia's articles are not medical advice, but are a widely used source of health information. For this reason it is vital that any biomedical and health information is based on reliable, third-party, published secondary sources and that it accurately reflects current knowledge."

This has the potential to change the sourcing policy from WP:RS to WP:MEDRS on a large number of Wikipedia pages, so please help us to arrive at a consensus on this issue. --Guy Macon (talk) 06:09, 3 November 2015 (UTC)

Relevant discussion

Please see Wikipedia:Reliable sources/Noticeboard#WikiJournal of Science. --Emir of Wikipedia (talk) 16:51, 3 December 2017 (UTC)

RfC

Please see Wikipedia:Reliable_sources/Noticeboard#RfC:Genetics_references Jytdog (talk) 17:02, 9 September 2018 (UTC)

Requested move 10 September 2018

The following is a closed discussion of a requested move. Please do not modify it. Subsequent comments should be made in a new section on the talk page. Editors desiring to contest the closing decision should consider a move review. No further edits should be made to this section.

The result of the move request was: moved. (page mover nac) Flooded with them hundreds 11:03, 20 September 2018 (UTC)

---

Wikipedia:Identifying reliable sources (natural sciences) → Wikipedia:Identifying reliable sources (science) – The original title of this page was Wikipedia:Identifying reliable sources (science-related articles). It was moved to the current title in 2010, apparently without any substantial discussion. The lead was later updated to explicitly state that the scope of the page was only the natural sciences. However, I don't see anything in the body of the page that is actually specific to natural science. The advice is just as applicable to social and formal science, or more generally any academic discipline that follows the contemporary model of scientific publishing (peer-reviewed journals, review articles, university press monographs, etc). – Joe (talk) 10:08, 10 September 2018 (UTC)

Support per the nominator's rationale. --Manduco (talk) 11:15, 10 September 2018 (UTC)

• Support per nominator's rationale. Also strip the mention that this is only for natural science. Some adaptation is necessary to cover social sciences, but far less than is implied. Carl Fredrik ^talk 21:08, 10 September 2018 (UTC)
• Support per nominator. The current title is confusing. Andrew³²⁷ 12:38, 17 September 2018 (UTC)

---

The above discussion is preserved as an archive of a requested move. Please do not modify it. Subsequent comments should be made in a new section on this talk page or in a move review. No further edits should be made to this section.

Note about primary genetics sources

An RfC at RSN has been closed, here. I have implemented it in this diff. Jytdog (talk) 01:04, 8 October 2018 (UTC)

Why isn't this official policy?

I've posted about it Wikiproject science, but no one's answered, so I'll just copy my statement over here in the hopes that someone sees it:

I was wondering why WP:SCIRS has never been elevated to the status of a content guidelines like WP:MEDRS has been? It seems that the suggestions in SCIRS are all perfectly reasonable and would prevent a lot of bad content from being added. I am thinking specificly of its limitations on human genetics sources: Wikipedia is currently plagued by horrible genetics section on human populations in numerous articles due to ignoring WP:SCIRS's requirement that only review articles be used. Different genetics studies often contradict each other and can be abused for various nationalist/other POV agendas. This is a particular problem for Central/East Asian articles, where various Pan-Turkic, Pan-Iranian, etc. editors war over which group various ancient peoples belonged to and add one-sided genetics sources to bolster their cases. There is currently a debate in which SCIRS is at the center at Talk:Uyghurs#No primary sources, for instance.

Anyway, WP:SCIRS seems like it should be a content guideline. How would we go about making it one? Or is there some problem with it I'm unaware of? At least two (good, non-problematic) editors at Talk:Uyghurs are deadset against implementing SCIRS, I know.--Ermenrich (talk) 13:22, 15 September 2019 (UTC)

@Ermenrich: I was also surprised to see it was not a guideline like MEDRS. I would also favor such a request. To make this a guideline, a broad consensus is needed, so I think something like a RfC (with a good enough participation - posting on the Teahouse might be useful) is needed. As I still have a relatively limited experience with english WP administrative procedures, I will refrain from initiating such a procedure, but I will gladly and strongly support it. --Signimu (talk) 20:49, 14 October 2019 (UTC)

@Ermenrich: Ah in fact after looking above, there was already an old RfC back in 2010 Wikipedia_talk:Identifying_reliable_sources_(science)#Promotion_to_guideline?. I think it would be worthwile to do another one, and then WP:CENT as suggested in this RfC if everything goes well. The main criticism was that it was a WP:CREEP mimicking WP:MEDRS, but I don't think so: of course they share similarities since they both rely on a hierarchy of sources based on the scientific method and its processes such as peer review, but WP:SCIRS strikes me as kind of a "light" version of WP:MEDRS where more latitude is being given to what reliable sources may be and what point of views to write about, which is adequate since scientific topics are not necessarily (and I would say seldom) as sensitive as medical and biomedical ones. Where in bio/medical topics, alternative views can mislead readers and patients with potentially disastrous consequences, in scientific topics these alternative views are much more dynamic and important for the scientific thought process. To give an example, between the superstring theory and the M-theory, although the former is much more accepted, the latter is also very interesting and well studied and should therefore be adequately represented, which with WP:MEDRS would probably not be the case due to concerns about the potential impact for human health (ok this is far fetched as there is no obvious human health impact but let's say there is one to make this point ;-) ). So IMO a "lightweight" version of MEDRS, consisting of a hierarchy of sources accounting for the scientific method but without the human health impacts precautions, would be extremely useful for the project to increase both the quality and trim needless discussions on WP:FRINGE theories. And so far, WP:SCIRS looks to me to have evolved to be exactly that now. --Signimu (talk) 21:05, 14 October 2019 (UTC)

The usual process is documented at WP:PROPOSAL (because of course a project that claims to be WP:NOTBURO has a policy on how to determine whether a page should be labeled a guideline...).

However: I recommend against making this proposal. Pages that get that magic "guideline" label at the top seem to cause editors to toss their brains out the window and approach articles with no latitude whatsoever – even if the guideline says, in large, blinking letters, that absolutely all editors are required to use judgment. I can't tell you how many times I've had to tell people that WP:MEDMOS's suggested section headings were just suggestions (it has always said exactly that at the top of the section that lists some popular section headings), or that WP:EL is about ==External links== and not about WP:CITEd sources. If you want people to use their best judgment, then, in practice, you do not actually want this page to be formally called a guideline. WhatamIdoing (talk) 21:13, 14 October 2019 (UTC)

Well, my main hope is that this could be used to clean up the awful genetics sections poluting Wikipedia's pages on human populations. There is a singular lack of editors using their judgment there, and if you point to SCIRS they usually just say it's not a guideline so they don't have to follow it.--Ermenrich (talk) 21:47, 14 October 2019 (UTC)

Have you made friends at WP:FTN and WT:MED? There are editors who are willing to help. WhatamIdoing (talk) 00:56, 18 October 2019 (UTC)

Contrary to comments before yours, I would absolutely recommend getting this to guideline status someday. It's all too often people reach for primary research articles first instead of secondary sources and wiki-lawyer to do so, including what you just mentioned or just making blanket statements that primary sources can be used per policy, therefore there is nothing wrong with their particular use.

That said, some work would need to be done here to give rough guidelines on when it's ok to use a primary source (and how) and when it's not. For example:

1. Primary research articles do have content in the introduction and to some degree the discussion/conclusions that can be used as secondary (though cherrypicking is a concern) when literature reviews are not available or the information is relatively uncontroversial (e.g., the life cycle of an insect species). Reporting of results within the primary research article should be avoided though until another source reports on those findings.
2. Stick to secondary sources (e.g., literature reviews, meta-analyses, etc.) in a controversial subject and avoid primary journal articles for content.

Those two alone already take care of the major problems with primary sources while avoiding concerns that have popped up if primary sources are excluded entirely from non-MEDRS areas. The main thing though is better describing how primary sources should be used before any promotion to guideline. Kingofaces43 (talk) 03:20, 18 October 2019 (UTC)

Thanks WhatamIdoing Agree with @Kingofaces43:, that's a nice proposition, and I also think that SCIRS should be more lenient on this part than MEDRS (else we should just use MEDRS ). Your proposition looks good to me, would you like to give it a try and modify the essay? Also I think we need this to be a guideline, judgment is always necessary and this will not block it, but it will help guide it and particularly on controversial topics, and in general will increase the overall quality of scientific articles. We need this if we want a serious encyclopedia! --Signimu (talk) 22:12, 20 October 2019 (UTC)

Signimu, I would oppose requiring review articles for issues other than human biomedical. I've seen this supposed requirement (there are editors who say it's already a requirement) used to engage in advocacy. We require it for MEDRS to err on the side of safety, but it means they lag, sometimes considerably. There is no need for us to do this with environmental issues, for example. If you're proposing to add that, it would need wiki-wide consultation, separately from any attempt to make this a guideline. SarahSV ^(talk) 22:49, 20 October 2019 (UTC)

SlimVirgin I see your point, then I would propose to simply stick with a simple general hierarchy of scientific information: primary studies are allowed (as they are reviewed, they should be considered reliable sources, depending also of course on other parameters such as journal etc), but reviews should be higher in terms of reliability and given more precedence, without necessarily removing primary studies as is done in MEDRS. I agree reviews might lag behind, but they are still more reliable than primary studies, where there is an inherent high variability. So it's nice to add cutting-edge info, but not for example for the LEDE. My main reason for supporting SCIRS is to have a guideline hierarchizing reliable sources for scientific articles. I think this is necessary for any domain, and more so for science. --Signimu (talk) 23:05, 20 October 2019 (UTC)

Here's my concrete suggestion: I changed «Primary sources should be used when discussing a particular result or recent research directions» to «may be used». I think this is clear enough and gives enough lattitude to editors judgment about whether or not to include primary sources. --Signimu (talk) 23:10, 20 October 2019 (UTC)

I made a few other edits I hope enhance the essay. I understand why it's not a guideline yet, it needs much more work and refinement to be ready. There are several wordings that were ambiguous and could be used to mean the opposite of what was intended (I fixed a few ones). A collaborative effort to make this essay RfC-ready would be great, and I would gladly participate! But I can't do this on my own. Maybe there's somewhere we can ask for help (with people experienced with guidelines drafting), someone has any idea? --Signimu (talk) 23:24, 20 October 2019 (UTC)

SlimVirgin, how do you feel about primary sources in genetics sections?--Ermenrich (talk) 23:40, 20 October 2019 (UTC)

Ermenrich, I can't comment on that specifically. The page advises against using them, and I have no problem with that advice. SarahSV ^(talk) 00:26, 21 October 2019 (UTC)

SlimVirgin I just saw you wrote several WP policies Would you be interesting in helping/supervising refining this into a potential policy? I can work on it, but I need guidance (but I learn fast ) --Signimu (talk) 02:11, 21 October 2019 (UTC)

Signimu, I'm sorry, but I don't think I'll have time.

The key is to make sure the page is not repetitive, that it contains no pointless sentences, that each word has a function, and that each sentence can stand alone and is not ambiguous. For example, you can remove the first sentence: "Wikipedia's science articles are not intended to provide formal instruction, but they are nonetheless an important and widely used resource."

You could tighten the second and third sentences: "Scientific information on Wikipedia should reflect the current state of knowledge. Ideal sources include reviews in independent, reliable published sources ..." With every sentence, imagine yourself in various editing scenarios and what the outcome of that sentence would be. Make sure you don't contradict the content policies. Go through the whole page in this way, and you'll have yourself a guideline. SarahSV ^(talk) 03:34, 21 October 2019 (UTC)

SlimVirgin No problem, thank you very much for the pointers, that's very helpful! I can see it's gonna take some time indeed I'm not sure I'll be able to refine everything, but I'll try to make it advance a bit at least, we'll see how it ends up --Signimu (talk) 03:38, 21 October 2019 (UTC)

SlimVirgin, can you provide specific examples of the accusation that sticking to secondary sources has been used for advocacy on any regular basis that would concern this discussion? Going to higher quality sources is typically a guard against advocacy, not the other way around, and policy already gives us a higher preference for secondary sources in terms of weight where concerns about advocacy would come up. That comment just seems to be going really out of left field in this discussion so far. Kingofaces43 (talk) 05:22, 21 October 2019 (UTC)