Wikipedia talk:Identifying reliable sources (medicine)

"Do not reject a high-quality study-type" ?

MEDRS says

Do not reject a high-quality study-type because of personal objections to: inclusion criteria, references, funding sources, or conclusions.

I know this has been discussed recently, and I may be being dim, but what does rejecting a study-type look like? Is it saying something like "I deny this is a even a systematic review because it is funded by cigarette manufacturers?" How can a "study-type" have "inclusion criteria, references, funding sources, or conclusions"? Those properties can only belong to instances of a type (i.e. an actual study), and not to a "type" itself. Alexbrn (talk) 06:02, 4 December 2015 (UTC)

Yes, "type" could be dropped. Johnbod (talk) 12:45, 4 December 2015 (UTC)

But then, is it actually right? In my experience the mentioned properties are often taken into account when evaluating sources. Alexbrn (talk) 12:53, 4 December 2015 (UTC)

I'm not sure the meaning is changed much. "Do not reject a study of a high-quality type because of ..." expresses what is presumably the intended meaning more grammatically. Johnbod (talk) 15:21, 7 December 2015 (UTC)

The "study-type" wording is new. It used to say "high-quality type of study". The word high-quality refers to the type of evidence (see: the entire rest of the section that this sentence is in), not to the overall reliability of an individual source. The main point to be made here is "No, you don't get to reject the meta analysis in favor of your cherry-picked randomized controlled trial, merely because you have a personal objection to some characteristic of the meta analysis". WhatamIdoing (talk) 01:54, 8 December 2015 (UTC)

Well we should go back, though "Do not reject a study of a high-quality type because of ..." is clearer, as it is an individual rejection that is being talked about. Johnbod (talk) 13:31, 8 December 2015 (UTC)

I see it as specifying that in this context, quality refers to its position on the hierarchy of evidence. "High-quality study" on its own could potentially refer to primary sources, and is also considerably more subjective. I don't think the objection being addressed is "I deny this is even a systematic review" as much as "This systematic review should be excluded/attributed because of my personal opinions about its inclusion criteria/references/etc," as opposed to objecting based on e.g. the weight of competing sources. This was discussed a few months ago in Archive 15 - it's the entire archive, but the second half is probably more relevant. Sunrise (talk) 02:30, 6 December 2015 (UTC)

WhatamIdoing recently proposed changing that wording to emphasize source quality over source type, since the way the language reads, we can't actually reject a meta-analysis that uses industry funded research that's been red-flagged as having possible conflict of interest. I thought that proposal might be a good idea, but still think there's issues with it, and yeah, we don't want to remove "type" because then we'll get primary studies. Now, the question is, if we have cigarette manufacturers funding studies that show cigarettes have no association with lung cancer and those get picked up in a meta-analysis, what do we do then? LesVegas (talk) 14:58, 7 December 2015 (UTC)

FTR: You have apparently not understood my proposal. WhatamIdoing (talk) 02:11, 8 December 2015 (UTC)

We describe the results of the meta-analysis, and we frame that within the suggestions of RS on the likelihood of bias. A good idea in general and it doesn't require us to rewrite any guidelines. Richard Keatinge (talk) 15:16, 7 December 2015 (UTC)

And we consider if such a deranged proposal is WP:DUE. There is more than sufficient evidence of the contrary that it is impossible that such a meta-analysis could be of acceptable quality. CFCF 💌 📧 15:20, 7 December 2015 (UTC) 

Due weight would indeed be important. I think that LesVegas is trying to make the point that a really high-quality study type should almost always be mentioned, however grave the suspicions surrounding it. Richard Keatinge (talk) 15:53, 7 December 2015 (UTC)

Despite the oft-repeated assertions that MEDRS requires us to use industry-funded sources, we most certainly can and do reject meta-analyses for all kinds of reasons – just not (a) in favor of weaker evidence and (b) due to personal objections to funding sources. We cheerfully reject bad meta-analyses in favor of good ones every day of the week, and we also reject sources that have been discredited by academics ("impersonal" objections), or carefully limit them to their proper WP:DUE weight. WhatamIdoing (talk) 02:11, 8 December 2015 (UTC)

These decisions are indeed a matter of weight and consensus. I would still start with a feeling that a large meta-analysis is generally worth mentioning in some context, and that serious doubts about its validity should form part of that context. Richard Keatinge (talk) 10:47, 8 December 2015 (UTC)

Yeah, I am inclined to agree with Richard Keatinge here that mentioning meta-analyses, systematic reviews and the like is usually a worthwhile endeavor and if there are noted issues with those based on funding, methods, or whatever, and we have criticisms which explicitly identify the junk sources, those criticisms should definitely be mentioned. If we have a meta-analysis which was funded by big tobacco, it's still notable even if it's garbage science. It's not for us as editors to question a published meta-analysis. It's for us as editors to just state what sources say, including critical sources. And there's plenty of criticism out there which are easily obtainable and can be easily included. Even the recently famous Coca-Cola study is worthy of inclusion here on WP, although probably 3-4 sentences of scientific rebuttals should follow. I wonder if we should include something in MEDRS saying just that, i.e., "however, if you include high quality study-types with questionable research due to funding sources, methods, etc, please ensure that criticisms also be included, should explicit criticisms of the source in question be available." Shouldn't we add something like that to make it more understandable? LesVegas (talk) 21:10, 8 December 2015 (UTC)

My point was that such an article will never be published in any respectable journal, and we can't act upon hypotheticals that are so unlikely that they will never come to pass. If such a study were ever published it is absolutely correct that editors question it, just not with personal objections. I don't find the current wording to be ideal, but those suggestions are simply superfluous, and don't address the issue at all – there will be a period in which there are no rebuttals for which that wording will actually cause a worse situation than what we have now. As has been mentioned before a WP:MEDDUE guideline is very much needed so that we can avoid this endless and mostly unproductive debate. CFCF 💌 📧 21:51, 8 December 2015 (UTC) 

CFCF that's simply not correct, there is an ample amount of research that has been published in very respectable journals and criticized thereafter. I have provided a lot of this very research in varying diffs during the RfC process. I may need to make a note of this in the current RfC, since it appears your opinion is based on not reading these diffs. So much research is industry-funded that researchers would have a hard time conducting meta-analyses if it was all government or independently funded. LesVegas (talk) 23:28, 8 December 2015 (UTC)

This is an entirely unrelated discussion about your example about cigarettes. You have multiple times raised extremely unlikely hypotheticals, but they are made no more likely just because you once provided some unrelated diffs. CFCF 💌 📧 20:40, 9 December 2015 (UTC)

A large meta-analysis might be worth mentioning in some context, but it might not be worth mentioning at all. To give a very relevant example, 167 meta-analyses about acupuncture have been published in the last five years, and even if merely a tenth of them are "large", then we cannot and should not mention all of them in Acupuncture. In that case, it's probably best to pick those that are most reputable and most representative of the overall literature. WhatamIdoing (talk) 04:36, 9 December 2015 (UTC)

Spot on WhatamIdoing. It's good to discern the possibility of a useful consensus. Richard Keatinge (talk) 20:17, 9 December 2015 (UTC)

Yes, agreed Richard, and I agree with WhatamIdoing there as well. Of course there's no way one could mention 167 meta-analyses in an article. I have, in the past, proposed something similar on the Acupuncture article, and am curious if we should add this into our guideline somewhere. Namely, what I proposed was that we mention claims for conditions that Cochrane says something about. So if Cochrane says, "Acupuncture is good for migraines" we can use meta-analyses for migraines. But when Cochrane says, "There's not enough evidence to suggest Acupuncture is efficacious for allergic rhinitis" we don't mention it, meta-analysis or not, or just mention a whole list of conditions which Cochrane says Acupuncture may or may not have benefit for. What seems to happen on that article is that editors will post every single meta-analysis that gets published, good and bad, and edit wars ensue on both sides. So not only is WhatamIdoing's suggestion practical, it also avoids edit wars on articles like that. And I really think something to this affect should be written into the MEDRS guideline. LesVegas (talk) 23:12, 9 December 2015 (UTC)

That is really more "MEDDUE" than "MEDRS", but it might be interesting to figure out how to address that. WhatamIdoing (talk) 06:01, 11 December 2015 (UTC)

You're really trying to interpret this in a way that furthers your point of view at every level LesVegas. Cochrane reviews are considered some of the most authoritative, and would be WP:DUE for exactly the example you say they aren't. CFCF 💌 📧 06:06, 11 December 2015 (UTC) 

WP:CGTW#8 keeps coming to mind. Alexbrn (talk) 06:21, 11 December 2015 (UTC)

CFCF, the Cochrane database has a lot of "no good evidence found" results in it. That doesn't mean that we need to go over to Cough syrup and say that there's no good evidence about whether cough syrup causes Alzheimer's. (Even though there isn't! I just ran a systematic review on the subject myself.  ;-) Unless a lot of sources are talking about whether cough syrup causes Alzheimer's, there's no need to mention the subject at all. By contrast, there are a lot of sources talking about acupuncture for allergic rhinitis (including approximately one review a year specifically on the subject of allergies and acupuncture, e.g., PMID 25269403), so in that instance, I would include it in the list of "sometimes used this way, but there's no good evidence either way" (or whatever the result of the studies is). WhatamIdoing (talk) 07:24, 11 December 2015 (UTC)

SO if this text boils down to "don't remove strong sources for capricious reasons" can't we say in more plainly? Or even remove it altogether as it's kind of obvious. Alexbrn (talk) 06:21, 11 December 2015 (UTC)

The text exists because we've had problems with this area. Also, nobody believes that their reasons for removing a strong source is capricious. WhatamIdoing (talk) 07:24, 11 December 2015 (UTC)

Then this should be dealt with by normal consensus-building. Sorry, but I really don't think that we need any change to MEDRS. If anything useful were likely to emerge from this mass of verbiage, I suspect we'd have seen it by now. Richard Keatinge (talk) 13:12, 11 December 2015 (UTC)

(Sorry I am arriving late at the discussion.) I have previously disputed the inclusion of this phrase with WAID here. Axl ¤ [Talk] 15:57, 20 January 2016 (UTC)

• I've changed the sentence to "Do not reject a source that is compliant with this guideline because of personal objections to inclusion criteria, references, funding sources or conclusions." SarahSV ^(talk) 02:54, 27 January 2016 (UTC)

Change to wording while the RfC is still running

I noticed this change. Thoughts? QuackGuru (talk) 16:39, 29 January 2016 (UTC)

Question on side

Browsing over the discussion, I kinda wondering why the formulation isn't simply:

• Do not reject a high-quality type of study due to personal objections

The problem with listing specific reasons without any qualifier is, that there probably always will be some folks claiming the list to be exhaustive and hence conclude objections against high quality studies based on grounds not listed are fair game.--Kmhkmh (talk) 18:58, 29 January 2016 (UTC)

Here is a constructive suggestion, thanks. Reducing bloat, rather than increasing it, must be a good thing. Personally I would consider removing even this sentence. Richard Keatinge (talk) 15:36, 30 January 2016 (UTC)

It's been brought up, more than once. On the whole, the dispute hasn't been about literal details like that, it's mostly about underlying intent. --Tsavage (talk) 15:51, 30 January 2016 (UTC)

Indeed. This accounts for its futile nature. Richard Keatinge (talk) 16:31, 30 January 2016 (UTC)

Guideline change: "Be careful of .. a journal that ... that reports material in a different field."

Wondering about this recent change] to the guideline, from:

Be careful of material published in a journal that lacks peer review or that reports material in a different field.

...to...

Be careful of material published in journals lacking peer review or which reports material mainly in other fields.

Doesn't changing "in a different field" to "mainly in other fields" open up the interpretation to in some cases include multidisciplinary journals like Science, Nature, or Scientific Reporst? It would seem that the intent of the guideline is to address cases where something is entirely out of place in a journal about something else, like a medical review in a political science journal, which is more clearly addressed in the previous wording. --Tsavage (talk) 13:12, 31 January 2016 (UTC)

No, the change does the exact opposite of what you're suggesting it does. It solidifies exactly what you mean it calls into question. This is silly. CFCF 💌 📧 13:25, 31 January 2016 (UTC) 

I think it's all superfluous anyway and we should scrap the sentence entirely. We already suggest that sources should come from journals that specialize in that topic when MEDRS mentions this under Biomedical Journals: or its content being outside the journal's normal scope (for instance, an article on the efficacy of a new cancer treatment in a psychiatric journal or the surgical techniques for hip replacement in a urology journal).LesVegas (talk) 04:13, 1 February 2016 (UTC)

Exactly. The more language, the more qualification of guidelines, the more they can be misused and need further qualification—it's a bueraucratic death spiral that makes it harder and harder to avoid long disputes focused more on specific PAG wording than on the content itself.

CFCF: "This is silly." Not silly: in this instance, in a recent glyphosate RfC, one of the recurring arguments against a source was that it was published in Scientific Reports (a high-ranked, MEDLINE-indexed, open-access, multidisciplinary journal from the Nature publishing group), which was challenged as somehow sketchy because of a kerfuffle over the principle of the journal's new paid fast-track peer review as a revenue source.

With that level of argument being commonly entertained by some regular subject editors, it is not hard to see a medical report in, say, Scientific Reports, objected to by arguing that the journal is not sufficiently qualified, citing that it "reports material mainly in other fields, per MEDRS," which SOUNDS different than "in a different field" - the former can make an argumentative case based on the predominant disciplines in the journal, the latter is more simply countered by, "Not applicable, it's a reputable multidisciplinary journal."

Or do we need more language to allow multidisciplinary journals, like Science and Nature? Or do we want to give them less weight in MEDRS?

If THAT seems argumentative and overly detailed in discussing guideline changes, look at the well-joined arguments over changes to MEDRS, or look at content arguments involving the guideline... The place to try to prevent various forms of rule-based gridlock and wikilawyering seems to be in the rules themselves. --Tsavage (talk) 14:43, 1 February 2016 (UTC)

Photographic example of PAGs gone wrong, [1]. Atsme^📞📧 17:11, 3 February 2016 (UTC)

I think an example might help, so here's one with which I am familiar.

Why Extreme Dilutions Reach Non-zero Asymptotes: A Nanoparticulate Hypothesis Based on Froth Flotation. Prashant S. Chikramane, Dhrubajyoti Kalita, Akkihebbal K. Suresh, Shantaram G. Kane, and Jayesh R. Bellare, Department of Chemical Engineering and ‡Department of Biosciences and Bioengineering, Indian Institute of Technology (IIT) Bombay, Adi Shankaracharya Marg, Powai, Mumbai −400076, Maharashtra, India, Langmuir, 2012, 28 (45), pp 15864–15875 DOI: 10.1021/la303477s

This is, in fact, a paper spruiking homeopathy. It made its way into a real journal because the peer reviewers failed to spot that it was a Trojan horse. It has subsequently been misused by homeopathy shills (e.g. Dana Ullman to propose that there is in fact a plausible mechanism by which homeopathy might work, and this was pretty clearly the intent. Iris Bell also has a long history of publishing pro-homeopathy "research" in a series of different journals, she gets away with it only once per journal, once the editors realise they have been taken for a ride, no further papers get in.

So the key question to ask is: are the editors and reviewers likely to be able to spot a ringer.

For what it's worth, I support the first version at the head of this section. A journal out of its field is vulnerable to being hoodwinked, a multidisciplinary journal should have the review processes in place to ensure it is not. The second wording would arguably put Nature in the same category as a subject-specific journal publishing out of its field, which I do not think is reasonable. In any case, we should only include primary sources where the results are uncontentious. A paper in a biomedical journal claiming to have found evidence for qi or meridians should not be included as a primary source because it's an extraordinary claim, we should rely on what secondary sources say about it, but a well-constructed study showing that reiki does not work and exposing an interesting facet of why people believe it anyway, would probably be fine since that finding is uncontroversial (other than to True Believers, who will never be satisfied with a reality-based treatment anyway). Guy (Help!) 09:09, 5 February 2016 (UTC)

On attempts to legislate Clue

Much of the debate above seems to be an argument between opposite entrenched camps determined to amend this guideline to either include or exclude research that favours their own personal views. However, when you read the actual proposals, all of it boils down to attempts to legislate Clue. What we should actually say is: use your good editorial judgment, and if in doubt seek consensus on the talk page.

The amended close proposed by Jamesday comes closest to this, IMO, but actually I think that any reference to exclusion by country is inappropriate. Any editor active in this area should be well aware that some countries almost never produce negative results. We cannot legislate reliability by changing this guideline, we should follow the sources which tell us. It appears on the face of it that people are trying to make rules that require us to ignore things they find ideologically inconvenient. A bit like North Carolina's attempts to wave seal level rises away. Guy (Help!) 09:46, 5 February 2016 (UTC)

"Some countries almost never produce negative results": Let me ask you, do you think using an 18 year old piece of research that includes primary studies that go back to 1966, none of which would make it onto our encyclopedia for obvious reasons, is actually a reasonable reason to exclude everything, including very high quality sources, from one particular country? Do you think it's reasonable to exclude US published research because it has authors that are Chinese? Not that it matters, in the RfC it was determined it wasn't reasonable. But I'm really curious about what you think. LesVegas (talk) 14:19, 5 February 2016 (UTC)

First of all, it's not just an "18 year old piece of research". A subsequent 2014 meta-analysis (conducted by Chinese researchers, and published in an alt-med journal) found that 99.8% of Chinese acupuncture trials reported a positive outcome and identified this as a major red flag. This is very much an ongoing issue in the literature, so minimizing it as ancient history is inaccurate. Separately, I don't understand the point you keep harping on about "primary studies". All meta-analyses look at "primary studies". How else would one detect bias in the literature, if not by looking at "primary studies"? Finally, I don't have any intention of using literature that is known to be biased in our Wikipedia articles, regardless of what the RfC shitshow was closed as this week. What I see is a small group of editors, most of whom have documented track records of poor-quality or tendentious editing on medical topics, trying to rewrite a core guideline to win a content dispute. I'm less charitable than Guy, in that respect. MastCell ^Talk 02:21, 6 February 2016 (UTC)

Recently described rare diseases

The guidelines say to avoid primary sources and to mainly use secondary sources. In general, this is good guidance. But what about recently described rare diseases? They might not have any secondary sources or they might have secondary sources that are not in-depth enough. Yet, they might have plenty of legitimate primary sources.

In particular, I am thinking about this article: Postorgasmic illness syndrome, and this exchange on its talk page: Talk:Postorgasmic illness syndrome#Deletion_of_a_recent_case_report. What do you think? --POIS22 (talk) 21:01, 12 February 2016 (UTC)

Update - close of original RfC on country of origin has been overturned

The close-review at AN has been closed. See close at the AN here, and importantly, the notes on overturning it in the archive here - the explanation is at the bottom. Please note the close not only overturns the original close, but also re-closes it, with the answer "no" with regard to adding "country of origin".

I suggest that LesVegas withdraw the RfC above about implementing the close, since it now has no basis. I also urge anyone unhappy with the close-review to carefully read the explanation I mentioned above. I will also note the overturning in the RfC above, in case LesVegas elects not to withdraw it. Jytdog (talk) 20:17, 14 February 2016 (UTC)

Still confusing secondary and primary sources

I've left this alone for about a year, but it's still not been fixed. The "Ideal sources for such content include: ... guidelines or position statements from national or international expert bodies" statement (emphasis added) is muddled thinking, and directly contradicts our sourcing policies. Organizational guidelines, from an actually authoritative institution, are likely reliable sources and can be treated as secondary or at worst tertiary, depending on their content and its source backing. But a position statement (a.k.a. a press release, advisory bulletin, etc.) is an almost invariably highly politicized primary source, reflecting the interests of industry players, regulatory bodies, medical professionals' associations, and other stakeholders with a viewpoint and with power and/or money in the balance. We should always treat these as primary sources, even if often high-quality ones; we must directly attribute or quote from them verbatim, just like we would for any other kind of organization's position statement or press release, and not repeat what they say in WP's own voice as if it were certain fact. The problem is immediately apparent when you compare the position statements of US vs. European organizations on things like legalization and treatment of addictive drugs, and those of bodies in different fields when it comes to topics like electronic cigarettes. There's a major difference between a medical organization's position statement on an issue that may have socio-political dimensions (when do they not?), and a guideline of actual best medical or medical-research practices from a scientific standpoint. This problem undermines the ability of people to take this guideline seriously. I've been a huge fan of its existence and continued improvement, but because of this really very basic but critical error, I cannot rely on MEDRS as guidance, and always return to our actual policies. If the "or position statements" wording is removed, or even just moved and discussed as primary sourcing, the problem goes away.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  22:41, 26 February 2016 (UTC)

No, you continue to misunderstand this. We mean major scientific and medical bodies. And position statements they put out summarize that body's perspective on whatever the subject of the position, which of course takes into account both the literature and mainstream practice. These are absolutely the kind of secondary sources we want. Now, I have seen people claim that some body is a major medical or scientific body, when it isn't one - the editorial board of the Lancet, was a recent example. About differences - major bodies may different with one another, just as reviews of the evidence may come to different conclusions. Jytdog (talk) 22:59, 26 February 2016 (UTC)

I'm not misunderstanding anything, Jytdog. It doesn't matter if your organization has 10 members or 10 mil.; a press release it issues giving its "official" internal position on a matter is a primary source. It's the very definition of a primary source. Just think about this critically for a minute, neutrally, and stop thinking in terms of who you respect and where your professional pride is, and so on. The NRA has millions of members. When it issues a position statement on something about firearms (even something technical, not about regulatory issues), or anything for that matter, it's a primary source. The Fraternal Order of Police (or insert any of dozens of other professional law enforcement associations) has at least hundreds of thousands of members; when it issues a press release, including about a central concern of law enforcement, and based on the professional input of its most senior members, it remains a primary source. The American Bar Association has millions of members; when it issues a position statement about a legal matter, even though the organization is composed of trained legal experts, it is, yes, a primary source. And so on. Would you like 10 more examples? 50? You have to drop this pretense that medical organizations are magically exempt from all skepticism about the press releases they issue, and that we must take on pure faith that what they say is unassailably neutral and represents a real-world consensus.

You're confusing the reputability of the author-publisher with the nature of the publication. You also seem to be mistaking "primary source" for "unreliable or unusable source". I could be the #1 foremost authority in the entire world on something, but what I publish on my blog as my opinion about it, without any editorial sanity-checking above me, it is by definition a primary source. It might be a very high quality one, and WP might use it – just like a press release from the AMA – but WP has to use it within the rules we have for primary sources, including attribution, not blind repetition of its claims in WP's own voice as facts. (That's not to say that some organizational releases contain no secondary material; as with even journal articles presenting new research, they often begin with focused literature reviews, but those lit. revs. do not cover the conclusion reached and stance taken, whether it is an organizational or individual one.

Let me present this as a question: In what way could it possibly harm our readers' interests for WP to attribute press release claims to the organization that issued the release? Can you come up with one single example of where identifying, in the prose, where we got the idea from is detrimental to our encyclopedic mission? That's all this change would require: Treat primary sources as primary.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  05:47, 27 February 2016 (UTC)

Please don't bring up all kinds of offtopic (and very loaded - heh) examples like the NRA. That is just not helpful. Your historigoraphical analysis is also missing the boat. The way you are doing that, any review is also primary for as to what the authors of the review came up with. And to that I can only say argh. Please hear this. When a major medical or scientific body puts out a statement or position paper, it is synthesizing the best evidence and practices that exist, in order to communicate about something that matters to the field in which they operate. It is not like an editorial someone writes about something that happens to be important to that individual. Please don't characterize them as press releases either. What is at stake here, is what we can express in WP's voice as "accepted knowledge." We rely on the best secondary sources for that, and positions/statements by MSMB (to abbreviate) are some of our best sources for that. I'll just add that your interpretation here is way outside the mainstream of the WP:MED community - I know you have been around a lot longer than me but you don't seem to be aware of how long this has been in here. Jytdog (talk) 05:57, 27 February 2016 (UTC)

The fact that you don't see how this situation is analogous to any situation where leading organization X in topic Y takes a primary stance on issue Z in their field, and think it's all "off topic" if it's not about medicine, is troubling, and is strong evidence in favor of my point that you're treating medical organizations as if they're magically exempt from our sourcing policies. But fine, I won't mention non-medical orgs again in the context, if you won't pretend I believe secondary sources are really primary ones; we all understand already what the difference is: Every literature review [that WP would ever consider using as a source, anyway] is subject to editorial review beyond its author. The nature of a press release or position statement is that its authors simply put it forth; no one outside the organization has editorial, peer-review control over what the AMA or FDA or whatever says before they say it.

I don't see that you're even understanding what the issue is, so I guess I'll just have to RfC this. But I hope that won't be necessary, given that there only seems to be one editor two editors anywhere declining to recognize the difference between publishing a review of research on the one hand, and, on the other, publishing a stand an entity is taking on the basis of that research. The key to understanding the difference is right there in your own wording: "synthesizing the best evidence and practices that exist, in order to communicate about something that matters to the field in which they operate." What is being communicated is what matters here; an official position statement on something is essentially a form of organizational activism; it is primary, necessarily, and is not simply a sifted summary of the research. If an organizational position statement were just a literature review, we'd call it a literature review not a position statement. Maybe it is actually possible to clearly distinguish between organizational publications that are purely secondary (or sometimes tertiary), and those with a stance-positioning component, but so far all I'm hearing from you is sheer denial that anything the AMA (or insert any other TLA of the medical establishment) says can be other than something we must treat as "accepted knowledge" (i.e. WP:TRUTH) and pass it on as such in our own voice. I'll repeat my question, since you didn't answer it:

In what way could it possibly harm our readers' interests for WP to attribute press release claims to the organization that issued the release? Can you come up with one single example of where identifying, in the prose, where we got the idea from is detrimental to our encyclopedic mission?
 — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  06:23, 27 February 2016 (UTC) Updated to account for comment below.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  07:07, 27 February 2016 (UTC)

It is pretty obvious that organization position statements are secondary and not primary. The organization did not directly conduct a clinical trial or perform other types of laboratory research to formulate a position and hence these position statements cannot be considered primary. Rather the organization weighed the available evidence which would include published studies. Hence by definition, these position statements are secondary (or tertiary). Boghog (talk) 06:38, 27 February 2016 (UTC)

Already addressed that. Repeat: "That's not to say that some organizational releases contain no secondary material; as with even journal articles presenting new research, they often begin with focused literature reviews, but those lit. revs. do not cover the conclusion reached and stance taken, whether it is an organizational or individual one." This isn't even a medical vs. other topics matter. If I were a journalist like Jared Diamond, whom everyone seems to like to pick on, and did a pretty thorough review of scholarship in a topic, my book is a secondary source for that scholarship review, but it's a primary source for novel conclusions I pull out of my own head based on that review, which are not connected directly to anything anyone else said before. This is the essence of the difference between primary and secondary sourcing. Even, as already noted twice, primary research papers frequently contain large sections of secondary material, but they are firmly in primary-land when they transition into their own data analysis and conclusions sections. There is no transmogrification wand possessed by the AMA and similar organizations that makes their politicized stance decisions, made upon their review of the research, into secondary material like the review is. Confusing a stance/position taken, predicated on a review of material, for the review of the material itself is like mistaking your personal fandom of the Dallas Cowboys [or whatever sportsball team], based on a review of their win/loss record, for their actual review of their win/loss record. It's a PoV identification of research [about research] with the feelings or decisions that research inspires. They are qualitatively different. General semantics addresses this frequently and concretely; it's Korzybski's classic fallacy of mistaking the map for the territory, the menu for the meal. Anyway, since neither of you will answer the basic question I keep asking, that indicates pretty strongly what an RfC on this should ask.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  07:02, 27 February 2016 (UTC)

Position statements are at least one step removed from an event (i.e., the clinical research on which the position statements are based) and hence per WP:PSTS, are by definition secondary. A RfC to confirm this inescapable interpretation of a long standing policy is a colossal waste of time. Boghog (talk) 07:31, 27 February 2016 (UTC)

Agree. With respect to content considering the evidence such position statements are at least secondary of course. The argumentation of the OP here (essentially, change this to my liking or I won't take it seriously) seems tendentious. Alexbrn (talk) 07:59, 27 February 2016 (UTC)

I think one of the key points is that a position statement (on a factual matter) in this context is not the same as it would be outside of science - instead of being a statement of opinion, it's a statement of analysis. If the AMA writes a statement to support/oppose a funding bill in Congress, I'd assume they didn't base that on empirical evidence drawn from their field of expertise, and therefore treat it as opinion. But when they include factual claims, e.g. "tobacco smoking causes lung cancer," those are based on synthesis of the scientific literature. And as usual, any individual statement could contain a mixture of fact and opinion. Sunrise (talk) 09:09, 27 February 2016 (UTC)

• @SMcCandlish: Thank you for your very well argued position on this. I have long wondered why MEDRS has made press releases from the AMA, FDA, WHO and various highly politicized entities that are strongly affected my money and political appointments, international trade and relations (and even interference from the chief executive) to be treated with the same respect and authority as secondary sources and reviews. These press releases by governing boards are completely different than commissioned reviews by those same bodies. Our article Medical consensus says this:

There are many ways of producing medical consensus, but the most usual way is to convene an independent panel of experts, either by a medical association or by a governmental authority.

I hope you do start an RfC on it. Can you cite specific cases or WP:PAG on Wikipedia that speak to the issue of press releases from well respected organizations requiring that they be subject to the kind of treatment or restrictions you are suggesting here? I think that would make your case all the firmer, rather than just mentioning the NRA or FOP. A better parallel would be how a press release from the Smithsonian regarding a matter of history would be treated (such as the 1995 decision on how to portray the dropping of the A-Bomb [2]). --David Tornheim (talk) 12:30, 27 February 2016 (UTC)

@David Tornheim: You are confusing position statements with press releases. They are not the same. Also by describing the AMA, FDA, WHO, ... as highly politicized entities, it is clear that you have an axe to grind. Boghog (talk) 13:40, 27 February 2016 (UTC)

Yes, and I sincerely hope we are spared an RfC on this—which would a waste of time, similar to that of having to read the verbiage above. If you can't be succinct, you can't demand others to take you seriously—and any position needing such repeating over and over again to be heard is doomed from the start. CFCF 💌 📧 14:19, 27 February 2016 (UTC)

Boghog, anyone who has actually worked for a large organization knows that AMA, FDA, WHO, etc., are highly politicized entities, by their very nature; it requires no "axe to grind" to observe this. The FDA and WHO are politicized by definition, being governmental/intergovernmental. And AMA is very obviously political, since it takes positions on national regulatory and public policy matters [3], [4], [5], [6], etc. The principal reason organizations like AMA and BMA exist is to represent professionals in their field, collectively, in public policy (i.e. politics, albeit mostly issue-based not electoral).  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  21:16, 27 February 2016 (UTC)

Resetting

I'd like to pull out what Sunrise said pretty clearly above, which I think clarifies SMcCandlish's underlying concern. MEDRS talks about "position statements published by major health organizations". The question, I think, is - can distinctions be usefully made among kinds of position statements (for example - statements that are meant to express "accepted knowledge" vs statements that express "judgements on social issues"), and if so, should MEDRS reflect those distinctions? Again what we would need to be able to do is meaningfully define those "kinds" of things, so that we can write guideline language. I think there is an interesting question there, that does have implications for the guideline. Can we first focus on defining kinds? Examples are helpful in this discussion. Jytdog (talk) 18:36, 27 February 2016 (UTC)

Thanks, Jytdog, for realigning the discussion more practically. The gist of at least one of these distinctions: Is it a medical science conclusion reached by experts convened by the organization, released without direct interference by the organization's brass, or it is a position being taken (or directly shaped) by the board/executives of the organization? This seems at first tangential to Sunrise's approach to the question, but they're actually closely related.

It takes only seconds to find examples of where AMA, BMA, and other organizations issue things that arouse controversy [7], [8], [9], etc. But not all controversies are equivalent; some of them are between organizations, between an organization and a regulatory body, between an organization and a vocal subset of its own members, between an organization and a political segment of the public, etc. And the nature of the controversy can vary, being scientific, political (including religious/moral), organizational (i.e. with regard to the organization's proper role/function), practical (theoretical vs. clinically applied), etc. There's probably a way to encapsulate the sort that affect the reliability issues. The line[s] between a statement of analysis and a position on public policy is not always clear, though. We see this in cases where similar organizations in different sectors of medicine, or in different geographical regions with different regulatory histories, come to conflicting conclusions and recommedations; they can't all be right at the same time, and the reason they're conflicting is obviously political in most such cases, not because one of them is collectively stupid and is doing bad science.

Maybe the line to draw, to start with, is simply whether there's a political component to the statement itself. I.e., any time an organization says something along the lines of "there oughtta be a law ..." they're clearly in press-release land not purely statement-of-science territory.

At any rate, I haven't come with some bone to pick over a list of cherry-picked politicized "smoking gun" statements I don't like; it's a general source classification matter from where I'm sitting. Some statements about the science are clearly being issued with the intent to affect public policy; maybe even most of them are. This is why I think it's safest to treat them as high-quality but primary sources and attribute the claims they make: "According to a 2016 American Pediatric Surgical Association statement, ...". It just seems the cleanest, and lowest-conflict approach. Virtually unassailable from a sourcing policy standpoint. Sunrises's "any individual statement could contain a mixture of fact and opinion" is the crux of the matter to me.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  21:16, 27 February 2016 (UTC)

The three links you provide are not helpful. Controversy over science doesn't mean the science itself is controversial. The morning after pill is safe and effective. Human activity is driving climate change. These are not scientifically controversial statements. Please proceed more carefully here. Our goal needs to be determining how to use statements by MSMBs to support statements of accepted knowledge.

Again, if you cannot have a concrete, careful discussion, this discussion is going nowhere. Jytdog (talk) 21:41, 27 February 2016 (UTC) (strike offtopic example Jytdog (talk) 22:02, 27 February 2016 (UTC))

• Agree with statement: "...treat them as high-quality but primary sources and attribute the claims they make: 'According to a 2016 American Pediatric Surgical Association statement, ...'. It just seems the cleanest, and lowest-conflict approach. Virtually unassailable from a sourcing policy standpoint." This is how I have seen them treated in other encyclopedias and in the RS. --David Tornheim (talk) 21:51, 27 February 2016 (UTC)

David that is abolutely incorrect. Wikipedia is not post-modern where everything is relative and the establishment is corrupt. Your continued efforts to make it that way will continue to fail. Jytdog (talk) 21:55, 27 February 2016 (UTC)

Jytdog: Please refrain from such ridiculous comments misrepresenting what I said. I never made either of the black/white assertions you are claiming. If you want to learn something about Postmodernism, try inserting "everything is relative" into the Postmodernism article (or some of the other things I mentioned on Jimbo's page regarding Hume, Plato, Kant and Descartes regarding epistemology that you mistakenly believe are Postmodernist) and see what happens. --David Tornheim (talk) 02:34, 28 February 2016 (UTC)

I should clarify that I don't mean that we need to say "According to the American Lung Association, the research indicates that cigarettes really do cause cancer." Where ALA is doing a review of the science and publishing their own essentially-a-literature-review, that's a secondary source. I meant that we should say "According to the American Lung Association, cigarettes are the leading public health risk for adults" or "According to the American Lung Association, cigarettes should be banned", or some other clearly political stance-taking. I think this is clearer in the formulation below. It would, technically, be "safe" to always attribute everything from organizational political statements, but it would probably be pedantic, and is unnecessary, when the material in question is scientific and secondary, and the claim made is factual not recommending changes in public policy.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  22:10, 27 February 2016 (UTC)

That is what I'm aiming for; I was not implying that every controversy that arises about these statements is one we need to be concerned about, but that there are qualitatively different sorts of them to be distinguished. But here's a completely different approach that could obviate any need to do such an analysis. Maybe something as simple as a one-word clarification plus a footnote will do it:
...scientific position statements published by major health organizations<ref>Organizational position statements are always [[WP:PSTS|primary sources]] for socio-political positions. Where they contain secondary material, such as a literature review, they are secondary sources for the claims in that material. In short, distinguish between a regulatory position about tobacco, which should be directly attributed, and a finding that the research demonstrates that cigarettes cause cancer, which may simply be cited.</ref>
 — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  21:48, 27 February 2016 (UTC)

That proposal is not bad to me. I struggle with the implication that the statement needs to explicitly review the literature to be considered secondary on the science/medicine - in my view it doesn't - but that is a workable proposal. Let's see what others have to say. Jytdog (talk) 22:00, 27 February 2016 (UTC)

Sure, there may be a better way to phrase it. Was just a first draft. I'm more concerned with walling out the political positioning than walling in the science.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  22:02, 27 February 2016 (UTC)

so maybe
...scientific or medical position statements published by major health organizations<ref>Organizational position statements are always [[WP:PSTS|primary sources]] for socio-political positions. Where they contain summaries of the evidence or describe consensus medical practice, they are ideal secondary sources for the claims in that material. In short, distinguish between a regulatory position about tobacco, which should be directly attributed, and a finding that the research demonstrates that cigarettes cause cancer, which may stated in Wikipedia's voice and cited.</ref> ? Jytdog (talk) 22:09, 27 February 2016 (UTC)

Works for me.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  22:11, 27 February 2016 (UTC)

SMcCandlish, what is at stake for you in this, btw. Was there one or more (hopefully more) specific content disputes where this issue came into play? That would people understand why this matters, beyond the abstact level. Thanks. Jytdog (talk) 22:59, 27 February 2016 (UTC)

I'm reluctant to drag up specifics, because several of them are contentious WP:AC/DS matters. The respondents so far in most of this discussion, both above and below this post, are mostly personally involved in two or more of these areas of disputation in which WP:ARBCOM lawyering has been invoked, so it's best not to dig into those sores in great detail, or the same personality conflicts are liable to pop back up. The short version is that socio-political statements (on top of medical/scientific analysis) by organizations are frequently used in contentious areas, and conflict tends to erupt when they are advanced in WP's own voice as facts instead of as the positions of the bodies/establishments advancing them. I've seen this at the electronic cigarettes articles, GMO-related articles, traditional Asian health vs. Western medicine disputes (TCM, accu*, yoga, etc., etc.), chiropractic and orthopedics-related articles (the latter in part because what that word means varies widely depending upon where you are in the world, and the former because there are, similarly, two widely divergent medical vs. newage hoo-ha approaches to it, without it being a difference of regional regulatory definitions), the idea of a scientific consensus that Race (human categorization) is a social construct and an alleged new medical consensus against the idea, etc., etc. I run into the problem pretty frequently, and it can take a long time to sort out in some cases because people are too often convinced that anything a big medical (or scientific-including-sometimes-medical) organization says must be true and can be reported as fact, even if it's clearly a public policy position, not a medical research analysis.

PS: From my position, I have no stake other than wanting to see no interpretational, gameable conflict between MEDRS and NOR/V/RS; I suppose from fringe promoters' PoV, I have an anti-fringe stake in it, but that's a pro-policy stake, so I won't apologize for it. >;-)  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  01:09, 28 February 2016 (UTC)

MEDRS amendment proposal

what do folks think about amending the definition of "Secondary" as follows, by adding the underlined parts?

and medical guidelines or position statements published by major health organizations.

References

- Thanks. Jytdog (talk) 22:30, 27 February 2016 (UTC) (note, removed proposed addition of "science or medical" before "position statements" per Sunrise below. Jytdog (talk) 04:02, 28 February 2016 (UTC))

• Strong Oppose—pointless, I try to hope anyone who reads this guideline will understand that it covers medical content without this clarification. Also, the "in short" section makes up over half the added material. I wrote an essay about this kind of nonsense: WP:VERBAGE. CFCF 💌 📧 22:47, 27 February 2016 (UTC) 
  • The "in short" means "in lieu of a detailed explanation, this simple example illustrates the point", but that was really pretty obvious, wasn't it? If you don't like "In short", replace it with "For example", or just remove it, and the passage still gets the point across. Since the "In short" is entirely optional, and your objection seems predicated on its presence, what's the point of the objection? PS: Your essay doesn't seem relevant, and is about talk page filibustering in the form of "proof by verbosity", unrelated to whether policy/guideline wording is clarified with specifics.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  01:55, 28 February 2016 (UTC)

I should have been more clear, but the essay was written with a specific person in mind — one who by sheer virtue of incoherence is able to change opinion in the middle of a discussion. I did not expressly mention filibustering, as I am uncertain that the conduct is intentional — but I am not the first to bring light to this "tendency to drown out discussions through sheer sesquipedalian verbosity"[10], even though I may be the first to write an essay in your honor. CFCF 💌 📧 10:02, 28 February 2016 (UTC) 

• I agree that the general idea here is correct, but I think the "scientific and medical" addition probably isn't necessary because the sentence already says that this specifically applies to biomedical information. This formulation could also be interpreted as classifying sources rather than content, when actually each statement (using that word in the generic sense) should be evaluated on its own. If the ALA writes a position statement with the goal of advocating for a cigarette regulation, but says "Tobacco smoke causes lung cancer" in the middle of it, then we can still treat that part as a summary of the evidence. I'd also want to define the boundaries in the footnote more clearly, maybe by linking WP:BIOMEDICAL, since socio-political positions are sometimes matters of fact which are scientifically uncontroversial. Sunrise (talk) 00:53, 28 February 2016 (UTC)
  • The distinction between being able to use the secondary "smoking causes cancer" data-based conclusion by just citing it, but needing to attribute the position that cig regulations are needed (more, at all, etc.), is precisely what I'm after here. I don't think it's harmful or redundant to clarify with "scientific or medical position statements", somewhere, even if we don't browbeat people with it repeatedly. What we have presently is a nutshell statement that reads "Ideal sources for biomedical material include ... position statements", which is by itself an overstatement, but perhaps not problematic if we address it later. The lead says "Ideal sources for such content include: ... position statements". The "such content" appears to refer to "biomedical information" but that's in a different paragraph. We repeat this a third time under #Definitions, with "A secondary source in medicine ... include[s] ... position statements", and that's where the error really is, an error of categorical overstatement that is "classifying sources rather than content", exactly as you put it. This is the place to fix it. It simply isn't the case that all positions statements by a particular type of organization are secondary sources or ideal sources for biomedical material/information, but we can probably just address the first of these (the "secondary sources" thing), and the second one (the "ideal biomed source" thing) logically follows. I also have to note that all this repetition, in three places almost back-to-back on the same page, is way more of a concision and redundancy problem than inserting a clarification at one of these places (the most explanatory one) that we mean a particular type of position statement. All the above "verbage"-related hand-wringing is misplaced. If "position statements" in this third instance were changed to "scientific or medical position statements", the point is made, and the issue is closed.

    That said, I'm having difficulties with "socio-political positions are sometimes matters of fact which are scientifically uncontroversial". While that may be the case in right-this-moment "ground truth" terms, with regard to a few particular topics, that sort of thing can change very rapidly and unpredictably based on tomorrow's data. We should still treat such positions as primary; there is no "cost" associated with doing this other than attribution. However, it's possible we're thinking of something different when we think of "socio-political positions"; I suspect this may be the case, because "socio-political positions ... which are scientifically uncontroversial" doesn't really parse well for me. It seems a bit like saying "legal positions that are philosophically uncontroversial" or "aesthetic positions which are ethically uncontroversial"; the first half does not strongly relate to the second, and there can be other valid controversies than the type identified in the second half. What I'm thinking of can be boiled down to this: Is the organization making the statement saying or implying a "should" in the public policy sphere, beyond an "address this problem" generality? If it's advocating or suggesting some kind of particular socio-political response, not just identifying a problem and suggesting that it be addressed somehow, nor suggesting particular responses in medical approaches (the actual purview of the organization in question), then it's not a medical research conclusion, it's a societal and political one. There's a latent OR concern that could come up if we try to approach a specific socio-political position as "scientifically uncontroversial"; it's a "says who?" problem. I'm having a hard time even thinking of one that qualifies, beyond mega-obvious generalities, like "smog is bad for people and the environment, so steps should be taken to reduce it."
     — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  01:55, 28 February 2016 (UTC)

• Support adding it there, without the "In short" wording (my bad!) which CFCF objects to, and we could drop the word "always" from the footnote to hopefully address Sunrise's concern. As noted above, we're saying the same general thing in three places at the top of the document, almost word-for-word; some of that could probably be compressed in the nutshell and lead, with the details being in the #Definitions section; that should address any wordiness concerns. Maybe the one in the lead could be reduced to something like "Ideal sources for such content are the specific types of secondary sources [[#Definitions|identified below]]. Primary sources should generally not ...." The nutshell could be trimmed from "... recognised standard textbooks by experts in a field, or medical guidelines and position statements from national or international expert bodies" to "recognised standard textbooks, or medical guidelines and position statements from major medical organisations". It's already covered in more detail under #Definitions, and the nutshell need not explain the obvious, that textbooks are standard and organisations major because they involve experts.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  02:10, 28 February 2016 (UTC)

One thing I am struggling with here is your statement in the subsection above, that you consider "cigarettes are the leading public health risk for adults" to a socio-political statement, that is primary and must be attributed. That's a science statement. Did you mean that? Jytdog (talk) 04:31, 28 February 2016 (UTC)

• Comment I think a little more work might be in order. The context of when the statement is made seems to me as important. I used the example of the Smithsonian's decision on the Enola Gay [11] to illustrate this. Even if the Smithsonian had used extensive literature reviews and secondary sources as part of any public statements about why they changed their exhibit, it would be difficult to categorize any of those statements by the Smithsonian at that time as NPOV in Wiki voice of science because the context of the decision was political rather than academic and independent. They would be unlikely to include the literature that contested their view with the same weight as literature that supported their view. Nonetheless, the sources they might have cited might be very usable. This is the reason many of us at the GMO articles have been [uncomfortable] with the use of Wiki-voice regarding AAAS [12] and AMA [13] [14] board claims about a "scientific consensus" on GMO safety that were made by these boards just prior to a vote the same year on California Prop 37 on GMO labeling. There was a lot of cherry-picking going on by the AAAS statement, which appeared to come straight from Monsanto's campaign materials, even though some of the materials cited by during the press releases are indeed good RS. So, I think SMcCandlish's language above needs to consider context of a statement and whether other major organizations have taken a different position. --David Tornheim (talk) 03:38, 28 February 2016 (UTC)

I have no idea what you are on about with the Smithsonian; it isn't pertinent to MEDRS and neither I nor anyone else is going to waste time figuring out if it is relevant. I have no idea what your point is here, with regard to the proposed change, except that you are making it clear that you are trying to change a guideline with broad applications to win a specific content dispute. So pretty much every argument you make going forward is going to be read in that light, and ignored. You still do not understand Wikipedia. You would do us all a favor if you kept your comment shorter as well. Jytdog (talk) 04:00, 28 February 2016 (UTC)

@Jytdog: Please stop with assumptions of bad faith. I will note that your allegation is in connection with GMO's for which you are topic banned. --David Tornheim (talk) 07:08, 28 February 2016 (UTC)

I have not breached my topic ban. That is the worst kind of wikilawyering. Like this other thing you are doing now - trying to manipulate a guideline to gain advantage in a content dispute that you (not me) are involved in, which could be anything. You are the one who introduced the specifics of your topic; not me, and I have not addressed them, and I will not. I am addressing the issues that SMcCandlish has raised, and addressing your transparent attempt to game the system. Jytdog (talk) 08:52, 28 February 2016 (UTC)

David Tornheim — to me is seems you are the one who assumed bad faith. Jytdog is not alone in judging your arguments to be bordering on irrelevant. CFCF 💌 📧 10:02, 28 February 2016 (UTC)