- Wiki's citation policy for biomedical articles is ridiculous I'm afraid. They don't allow (OK maybe not quite, but close enough) you to cite actual scientific papers. Somehow the actual primary source is bad. Nobody has actually been able to tell me how that even remotely makes sense. It just means our articles are always a few years behind reality. As to the rest of your question, the only source I could find in a quick search is here. Fgf10 (talk) 20:01, 21 January 2021 (UTC)[reply]
- "actual scientific papers" that are invariably garbage primary sources. Abductive (reasoning) 07:18, 22 January 2021 (UTC)[reply]
- All of science is garbage? Is that you, mr Trump? Fgf10 (talk) 10:24, 22 January 2021 (UTC)[reply]
- Many reported results prove, on further examination, not to be reproducible; jumping on any new result will result in a lot of non-information being included. --Lambiam 11:48, 22 January 2021 (UTC)[reply]
- And conversely I actually did find a ref for "Insufficient evidence to rate effectiveness for" various intestinal problems.(Medline) As others note, Wikipedia is by policy conservative with reporting. WP:MEDRS is the consensus guideline for medical topics. We actually have an article specifically about the replication crisis that affects medical research. We (WP editors) are simply not qualified to make pronouncements about what primary research is valid vs flawed. DMacks (talk) 12:25, 22 January 2021 (UTC)[reply]
- As a followup to Fgf10's PMID:20738174 2011 primary-research (single trial) ref and Abductive's hinting that the actual researchers might not have a neutral perspective or well-designed/validated work, PMID:33149763 is a 2020 very recent meta-analysis that cites it. The first ref sounds promising on its own ("The results validate in vivo and in vitro antidiarrheal activity of Berberis aristata extracts and provide its chemical fingerprint."). But it takes the independent review for us to learn that "The quality of evidence of included trials was moderate to low or very low": even with a total of 38 trials, there are indications of effectiveness but "there is still a lack of high-quality evidence for evaluating the efficacy and safety of berberine." DMacks (talk) 12:33, 22 January 2021 (UTC)[reply]
- There are SEVERAL VERY GOOD reasons why we need more than a single primary source paper to satisfy Wikipedia's requirements:
- Primary source papers, like publications in scientific journals, provide us with raw information and data, but do not often place that data and information into context. They don't provide us with any means of analyzing the research to give it any significance. Is it important or insignificant? Does it change our understanding of a particular subject or not? Does it even matter? Without that information, we can't say any of that in the Wikipedia articles. Primary source papers, used in conjunction with secondary sources that provide analysis and context, are absolutely welcome in articles, but without secondary sources to provide analysis and context, by themselves are not useful.
- Many research papers are published on small studies which may or may not have enough data to draw firm conclusions from. Many, if not most, such papers are published only to show that a subject is worthy of further exploration and not that the matter is settled. So a publication may be done that shows a preliminary study shows a certain level of effectiveness on a small sampling of patients on a limited time scale. The purpose of that paper is only to show that it would be worth the time and effort to scale up the study and do more research. These kinds of papers form the bulk of peer-reviewed journals, and by themselves are not sufficient to draw definitive conclusions on things, especially with regard to medications and their effectiveness.
- The subset of "all journals" occupied by "well-regarded journals" is small; JAMA is of a different calibre than many fly-by-night scientific presses. Much of the scientific journal world is driven by supply-and-demand, and given the publish or perish mentality in academia, that tends to mean that there's a LOT more published than is strictly useful for our purposes. Which is not to say that such journals do not have a peer-review process, but it does tend to overwhelm the system with research of dubious import. It takes both time and good system of external analysis to weed out the wheat from the chaff, and at Wikipedia That's not our job. We need to wait for secondary sources to do that for us.
- That's why MEDRS and other guidelines at Wikipedia are written the way they are. --Jayron32 15:07, 22 January 2021 (UTC)[reply]
- I've published many papers, so that's hardly new to me, thank you. None of those are valid reasons, you just use decent sources. Jesus, it's not hard. It's exactly the same as with any other sourcing. There's a reason our biomedical articles are such a joke. EDIT: Also, secondary sources introduce errors, there have been numerous occasions where reviews have cited my work for a statement that is either not in the paper they cited or says the exact opposite. Therefore you should always go the primary source, not someone's potentially flawed interpretation of it. Fgf10 (talk) 16:12, 22 January 2021 (UTC)[reply]
- Almost any fringe science claim you can think of has at least one plausible-looking paper in a real, respected journal, and the true believers will gleefully post it as holy writ. Sure, scientists in the field will roll their eyes and say "I'll bet that's nonsense!", but that's a subjective call that's not really compatible with a collaborative encyclopedia anyone can edit.
- Nobody seems to be able to make up a rules-based way of distinguishing between "New discovery that is probably real", "New discovery that will probably vanish under closer scrutiny", and "Total nonsense that is probably from a biased researcher". Your suggestion to "Just use decent sources" implies that we already know which is which. And maybe a researcher in the field would intuitively, but that's not what Wikipedia needs. ApLundell (talk) 18:48, 22 January 2021 (UTC)[reply]
|