Wikipedia:No original research/Noticeboard: Difference between revisions

Welcome to the no original research noticeboard
This page is for requesting input on possible original research. Ask for advice here regarding material that might be original research or original synthesis. Include links to the relevant article(s). Make an attempt to familiarize yourself with the no original research policy before reporting issues here. You can also post here if you are unsure whether the content is considered original research. Sections older than 28 days archived by MiszaBot II. (For help, see Wikipedia:Purge) If you mention specific editors, please notify them. You may use {{subst:NORN-notice}} to do so.
Additional notes: "Original research" includes unpublished facts, arguments, speculation, and ideas; and any unpublished analysis or synthesis of published material that serves to advance a position. Such content is prohibited on Wikipedia. For volunteers wishing to mark a discussion resolved, use {{Resolved|Your reason here ~~~~}} at the top of the section.
To start a new request, enter a name (section header) for your request below:

Is it original research to take a table from an article on genetics and interpret it?

For about two weeks User:Tursclan and I have had a slow moving argument at Etruscan civilization (edit | talk | history | protect | delete | links | watch | logs | views) and Latins (Italic tribe) (edit | talk | history | protect | delete | links | watch | logs | views) with User:LambdofGod and their earlier IP User:185.218.35.216 over the use of (FigS29) which can be seen on p70 here. An example of its use is at Latins (Italic tribe)#Genetics which is currently fully protected by User:Swarm. I suspect there is cherry-picking from the article in the entire paragraph, and I don't think we should say "A 2019 genetic study by Stanford", but my main problem is with this sentence:

"In addition, genetic analysis (FigS29) shows that the Iron Age population had a much lower frequencies of SNPs associated with both light skin and light eye pigmentation compared to modern Italians, who instead are similar to other modern Europeans (British, Finnish and Spanish), although the authours are cautious about these results

Note that the quote isn't from the article but, like FigS29 from the supplementary materials [ here] although it's sourced to the article.

LambdofGod's last reinsertion is with the edit summary "Considering that neither Tursclan or Dougweller responded to my talk page, It's obvious that you guys have no argument beside "it hurts my feeling"" despite my posts at User talk:LambdofGod and User talk:Doug Weller#Etruscan civilization where I try to explain policy and they reply " I am exposing a fact, not my personal opinion, and the data are autoevident for anyone with a minimum of genetic knowledge. You can't really expect the authours to write a paragraph for all their 40-50 images." See also Talk:Etruscan origins#Genetic studies: recent edits on SNPs associated with light skin and blue eye where Tursclan has tried to explain to him. Doug Weller talk 15:55, 25 December 2019 (UTC)

More reinsertions, the latest being "In addition, genetic analysis (FigS29) shows a massive increase in the frequencies of SNPs associated with light eyes and light skin from the Iron Age to Medieval and Early modern population period, with the modern population resembling other modern European populations." with an edit summary "Fig. S29. Allele frequencies for alleles of functional importance. Imputed genotypes for alleles previously shown to be of functional importance (and under selection), denoted by putative function, associated gene, variant ID, and derived allele, are shown for study individuals from central Italy, ordered by time on the x-axis. For reference, the population allele frequency for three present-day populations in the 1000 Genomes Project (British/GBR, Finnish/FIN, Spanish/IBS) are designated by" which is an excerpt from the only comment on S29 I can see, "Allele frequencies for alleles of functional importance. Imputed genotypes for alleles previously shown to be of functional importance (and under selection), denoted by putative function, associated gene, variant ID, and derived allele, are shown for study individuals from central Italy, ordered by time on the x-axis. For reference, the population allele frequency for three present-day populations in the 1000 Genomes Project (British/GBR, Finnish/FIN, Spanish/IBS) are designated by the first letter of the population name. Sample points for study individuals are colored by their time period. For each variant, a LOESS (locally weighted smoothing line) is plotted all points excluding the three modern populations." Again this appears to be OR. I wish we had a policy on genetic peer reviewed articles not allowing us to use material not in the abstract (and that only if written by the authors) or the concluding discussion. Doug Weller talk 11:44, 26 December 2019 (UTC)

Yes - I think best practice is to use only quotes from the source for conclusions. Johnbod (talk) 12:09, 26 December 2019 (UTC)

References

• Left a note at WT:MED. GMG^talk 11:51, 26 December 2019 (UTC)
• I saw the note at WT:MED. Doug, do you have a genuine concern about the fundamental factual claim being incorrect, i.e., you think that the skin color likely didn't change during those ~25 centuries, or that skin colors might have gotten darker during that time? I haven't looked at the source, but I'm pretty sure that if you wrote to the authors and asked them if that's what their paper meant, when brought down to the simplest words, that they would add some hedging (because what if their 1000 genomes aren't truly representative?) but agree that it was accurate summary of the relevant material actually in the source (and therefore not a case of original research). The problem for Wikipedia (and the reason that I haven't bothered to look at the source) is that including information from any single recent primary source isn't WP:DUE for these articles. That content isn't important enough to mention in such general articles unless and until it's picked up by a secondary source, such as a textbook. WhatamIdoing (talk) 16:55, 26 December 2019 (UTC)

This the actual description from the article.

Fig. S29. Allele frequencies for alleles of functional importance. Imputed genotypes for alleles previously shown to be of functional importance (and under selection), denoted by putative function, associated gene, variant ID, and derived allele, are shown for study individuals from central Italy, ordered by time on the x-axis. For reference, the population allele frequency for three present-day populations in the 1000 Genomes Project (British/GBR, Finnish/FIN, Spanish/IBS) are designated by the first letter of the population name. Sample points for study individuals are colored by their time period. For each variant, a LOESS (locally weighted smoothing line) is plotted all points excluding the three modern populations.

Pay attention to the last part. They used an algorithm to show graphically the change over time of frequencies of certain SNPs.

I'm awaiting for your response. LambdofGod (talk) 21:29, 26 December 2019 (UTC)

"That content isn't important enough to mention in such general articles unless and until it's picked up by a secondary source, such as a textbook"

The content is shown in the fig29 of the original article so it's important enough to be mentioned. LambdofGod (talk) 23:51, 26 December 2019 (UTC)

That is wrong and is not how Wikipedia works—anyone can edit does not mean that anyone can cherry-pick factoids and use them to slant an article in a particular manner. As explained above, a secondary source would be required to conclude what Fig. S29 represents, and to demonstrate that the conclusion is WP:DUE for the article. The caption quoted above does not mention "skin" or "eye" or "pigmentation" so concluding that the figure proves certain findings regarding those matters is classic original research and is not permitted. Johnuniq (talk) 06:33, 27 December 2019 (UTC)

agree w/ Johnuniq--Ozzie10aaaa (talk) 22:05, 28 December 2019 (UTC)

"The caption quoted above does not mention "skin" or "eye" or "pigmentation" so concluding that the figure proves certain findings regarding those matters is classic original research and is not permitted."

Skin and eye are mentioned in the actual graph (fig29), so it's proven. There is no slander at all considering that now all archeogenetic papers includes analysis about functional SNPs and the results are often commented. Case in point this is from the article about the Yamnaya culture:

"Physical characteristics The genetic basis of a number of physical features of the Yamnaya people were ascertained by the ancient DNA studies conducted by Haak et al. (2015), Wilde et al. (2014) and Mathieson et al. (2015): they were genetically tall (phenotypic height is determined by both genetics and environmental factors), overwhelmingly dark-eyed (brown), dark-haired and had a skin colour that was moderately light, though somewhat darker than that of the average modern European.[28][7] Despite their pastoral lifestyle, there was little evidence of lactase persistence.[6]" LambdofGod (talk) 08:33, 27 December 2019 (UTC)

A caption at least has a chance of being a summary of the author's view (which still would not overcome the primary source problem). However it is original research (and UNDUE) for an editor to interpret a graph. You might browse WP:INDENT. Johnuniq (talk) 08:49, 27 December 2019 (UTC)

I have looked at the graphs shown on Fig S29. I have several concerns about Lambdofgod's interpretation.

My first concern is that, as far as I can tell, each dot on the graphs represents a single "study individual", i.e. one person. The sample size is small.

The authors do not include R-squared values to establish the quality of fit of the regression curves. This makes it difficult to know how well-modelled the regression curves are.

No statistical analysis is provided to show if allele frequency is significantly different between the different age-based populations. I suspect that the reason for this absence is that the differences are not statistically significant. This is also related to the small sample size.

As humans, we have evolved to be very good at pattern recognition—to the point where we see patterns even where patterns don't exist. Even though it looks like there may be a trend, can we be sure that a trend really exists? This is why the semi-objective techniques of statistical analysis are so helpful.

My interpretation: these findings might imply a trend towards lighter skin in more modern populations, but this requires further analysis of much larger sample sizes before we could declare this with any certainty. Of the genes tested, SLC24A5 looks like the most promising. In any case, our interpretations of these graphs should not be included in Wikipedia's article. Axl ¤ [Talk] 12:06, 2 January 2020 (UTC)

occupational stress

I am looking for advice on the occupational stress article. There is a large heavily weighted section in the lead which is not even discussed in the actual article itself. It has no real relevance to the article and seems promotional. It appears to be original research. Would appreciate other's opinions. Lightningstrikers (talk) 13:21, 10 January 2020 (UTC)

Yes, I agree, that second paragraph needs to not be in the lead. You can change it yourself if you like.--Epiphyllumlover (talk) 01:48, 11 January 2020 (UTC)

SYNTH, NPOV

Steele dossier - I'm requesting input regarding what appears to me to be a classic case of noncompliance with WP:NOR (SYNTH), and WP:NPOV. I am also of the mind that if one issue is resolved, the other with possibly self-correct. I'm going to focus on a single paragraph from a rather lengthy and detailed lead in a topic area I just know all editors and admins love to edit. You can thank me later. 😎

Contrary to a conspiracy theory pushed by Trump, Fox News, and many of Trump's congressional supporters, the dossier was not the trigger for the opening of the FBI's "Crossfire Hurricane" counterintelligence investigation into Russian interference in the 2016 presidential election campaign. It did play a central role in the seeking of FISA warrants on Carter Page in terms of establishing FISA's low bar for probable cause.

I realize we can state several facts in a single sentence citing different sources as long as we don't reach or imply a conclusion not explicitly stated by any of the sources; however, the various sources that were cited in that paragraph were used to not only form an absolute conclusion but to justify stating it in WikiVoice, which is not only SYNTH, it is noncompliant with NPOV.

The CBS News report that was cited for "probable cause" in the last sentence of the above paragraph also states: "However, the Horowitz report is not the final word on the origins of the investigation. U.S. Attorney John Durham is leading a separate review of the FBI's investigation, and after Horowitz released his findings, Durham also questioned the conclusions." There is no mention of this important fact. It is also a known fact that the IG is limited in both scope and reach outside the department which the IG report and Horowitz himself admitted - again, no mention. Durham's probe is a criminal investigation, and it includes information from outside the Justice Department, to include testimony from witnesses outside the US. There is also the AP report published by PBS News Hour that corroborates the information, and like the CBS report, is neutral and presents all relevant sides, which is what WP articles are supposed to do.

Sources

1. Is it SYNTH?
2. Is it compliant with NPOV?

Protests of 2019

Protests of 2019 was controversial almost immediately after its creation. I was not its creator, but I felt that the sources tended to support the creator's claim, so I developed the article. The final result of the deletion discussion, mainly about concerns for OR/SYNTHESIS, was no consensus and the article was kept. Subsequent mainstream media a month later seems to support the existence of the topic as a valid encyclopedic topic.

In the last week or so, a new user who is clearly quite enthusiastic and is making a good effort to provide sourced material has been adding quite substantial content. Since I've contributed most of the material to the article, I'm in a highly non-neutral position for judging what should or should not be in the article; the article is not mine. However, the new user does not seem to realise that what s/he is contributing is really pushing the limits of OR/SYNTHESIS in an article that started off with these concerns, and s/he does not understand that the talk page is the place to respond with concrete counterarguments, and that edit summaries are not sufficient for sorting out a serious content disagreement. The Protests of 2019 article is presently just two clicks from the main page (click (1) In the News/protests; click (2) infobox - Protests of 2019 at the top), so the OR could risk becoming embarrassing - or result in a new deletion proposal because of the added OR material - if there are no other editors participating and explaining to the new user. New users have the right to learn, and it's understandable that the new user is suspicious about listening to the opinion of just one person. Boud (talk) 18:36, 16 January 2020 (UTC) (minor copyedit Boud (talk) 18:40, 16 January 2020 (UTC))

The new editor has finally responded on the talk page, but by creating a new section rather than responding in specific talk page sections on specific editing issues; and in parallel has done a wholesale revert. However, s/he does seem to have made an effort on the Common causes section, without (in my judgment) quite getting the idea of OR in this context. Help in recommending that the editor focus on specific talk page sections concerning specific issues to work step by step, and independent opinions on these specific issues, should not be difficult for people who watch this noticeboard. The two specific issues/talk page sections are include Brexit? and include immigration/xenophobia paragraph?. Boud (talk) 11:59, 17 January 2020 (UTC)

Interpreting a source

For several weeks now, members of WikiProject Motorsport have been debating the meaning of the Sporting Regulations for the World Rally Championship. These are the rules produced by the governing body of motorsport, the FIA. The passage in question is Article 26, which reads as follows:

26. SEASONALLY ALLOCATED COMPETITION NUMBERS

26.1 MANUFACTURERS

P1 drivers may request a specific number provided that the application is endorsed by the FIA and the Promoter. Number 1 may only be chosen by the World Champion driver of the previous season. Requested numbers may not be greater than 99.

26.2 OTHER DRIVERS

Competition numbers shall be allocated rally by rally, according to the provisional classification of the Championships concerned.

It is this idea of "seasonally allocated competition numbers" that is proving problematic. One of the editors involved, Tvx1, put forward the following interpretation of Article 26:

Having taken another look at the sources in the article, as well as at the sporting regulations, I'm no longer convinced that these drivers/crews have chosen career numbers. Neither the sources, nor the regulations mention "career numbers". They all actually talk about season/seasonal numbers. It seems like they only reserve a number for the duration of a season. While it is likely that crews will pick the same numbers over multiple seasons, we can't really be certain of that.

His argument is that because the section is referred to as "seasonally allocated" numbers, that means a driver chooses a number for one year at which point they need to reapply for that number. I believe this to be original research on his part for the following reasons:

• The numbers are only referred to as "seasonally allocated" in the title of Article 26. The body of Article 26 refers to "permanent numbers", and the phrase "seasonally allocated numbers" is never used again.
• I have searched several times for sources that support Tvx1's claim. When I use a search term for wrc "seasonal numbers", the only hits I get related to the subject are the WT:MOTOR discussion where Tvx1 made his claim.
• There are a variety of sources out there detailing the number system—the rule was first introduced in 2019—which use the term "permanent numbers". These include the FIA website, wrc.com, Autosport and Speedcafe. These are four of the most reliable publications, which are routinely used across the scope of WP:MOTOR.

Furthermore, the FIA website details the specific changes to the Sporting Regulations year on year. This passage outlines the nature of Article 26:

In order to give consistent identity to drivers and assist with promotion, Priority 1 drivers will be free to choose their permanent car number from 2019, except number 1, which will always be reserved for the reigning World Rally Champion.

This specifically refers to "permanent numbers" rather than "seasonal numbers". I have requested that both Tvx1 and Pelmeen10—who supported his interpretation—share any sources that they have to substatiate the "seasonal numbers" argument. They have either refused, ignored the request, or claimed that the burden rests with those who disagree with them.

In the past week, a new source has become available: the entry list for the first round of 2020. It shows that two drivers (Sébastien Loeb and Takamoto Katsuta) are competing with different numbers to the ones they used in 2019. This source was not available at the time Tvx1 made his claim. Tvx1 is claiming that this proves him right; however, the entry list only shows that the numbers have changed. It does not explain how the number changes came about. I believe this to be synthesis.

There are three things that I would like to see happen in this discussion:

1. I would like Tvx1 and Pelmeen10 to share whatever sources they have to substantiate Tvx1's original claim.
2. I would like members of this noticeboard to evaluate all of the sources presented and determine whether or not Tvx1 and Pelmeen10 have engaged in original research.
3. I would like members of this noticeboard to offer some idea of how to handle the paradox created by the Monte Carlo entry list for future reference.

Finally, I know that this seems like a very minor thing to come to a noticeboard for. However, the discussion is taking place at WikiProject Motorsport (rather than WikiProject World Rally) and so has the potential to affect every single article within the scope of WP:MOTOR. Mclarenfan17 (talk) 10:59, 18 January 2020 (UTC)

Let me start with a disclaimer. I am an on-off contributor to the discussion in question.

I have to agree with Mclarenfan17 that when Tvx1 came forward with the claim originally it constituted original research. I also think that Tvx1's claim that drivers must reapply for numbers every year is still original research. However I have to agree with Tvx1 that the Monaco entry list shows that driver numbers can change, this is supported by this source: [1]. How or why these number changes came about is irrelevant and this source shows that numbers can change (I.e. they are not permanent) - this means that it would be original research to assume the numbers would stay the same (as explained to Mclarenfan17 in the WT:MOTOR discussion).

To conclude up until the release of the Monaco entry list I believe that Tvx1 and Pelmeen did engage in original research but they're not anymore as more sources have come to light. And articles should reflect the most recent sources when various sources contradict each other (ie. The articles should list numbers as TBA or TBC until an entry list comes out for that season).
SSSB (talk) 12:10, 18 January 2020 (UTC)

(edit conflict)SSSB, can you actually point to a diff of a comment where I (or Pelmeen10 specifically stated that I/they think that crews have to physically reapply for the their numbers at the end of each season? And how does it even matter? As you point out yourself, how or why the number changes come about is irrelevant. The only point I and Pelmeen10 ever made was that there was no evidence whatsoever that these numbers were fixed for their entire careers, which is what I have reiterated time and time again during that WT:MOTOR discussion. Even before the Monte Carlo entry list was published you stated repeatedly that there was insufficient evidence for Mclarenfan17's claims and that listing the numbers as TBA was a sensible way forward. What's wrong with being prudent? In the end, because of Mclarenfan17's antics we ended up listing incorrect numbers for a couple of crews on that article for weeks. I would like to know why we are being accused of engaging in OR, when actually Mclarenfan17 actually did so even more clearly by claiming and insisting that crews' numbers are fixed for their careers without ever providing any concrete evidence for that claim, forcing us to actually have incorrect information in an article for weeks. Even now with clear evidence that these numbers aren't fixed for entire careers and without any support in the WP:MOTOR discussion, within the last 24 hours [2][3], they kept reinstating crew numbers that have not been announced for 2020 yet in the 2020 article on the basis that that "they stay the same as last season".Tvx1 18:12, 18 January 2020 (UTC)

I believe that Tvx1 and Pelmeen did engage in original research but they're not anymore as more sources have come to light.

That does not justify what they did in the first place. They had no way of knowing that source would become available when they made that claim and their ongoing refusal to provide sources amounted to disruptive editing. It's especially galling considering that they insisted others produce sourcses.

Furthermore, the Monte Carlo entry list does not actually prove the original claim to be correct because the source does not offer any context. Tvx1 specifically claimed that every driver would have to go through the process of reapplying for a number, but the entry list only demonstrates that some numbers have changed. Two potential scenarios emerge:

1. Every driver had to reapply for their number. Where most kept the same number, some changed.
2. Every driver had a permanent number, but some decided to change and applied separately.

The Monte Carlo entry list does not establish either scenario as having happened, so Tvx1 and Pelmeen10 cannot claim to be right. They're quick to point out the two drivers that changed, but they're ignoring the six that did not. They should still show the sources that they used to justify the original claim that they made in November, or at least admit that they never had a source, because this sort of behaviour should not be overlooked. Mclarenfan17 (talk) 18:07, 18 January 2020 (UTC)

Please provide a diff were I made that specific claim? And exactly what justifies what you did? What evidence did you ever provide that supported your claim that these numbers were supposedly fixed for their entire careers??Tvx1 18:12, 18 January 2020 (UTC)

Please provide a diff were I made that specific claim?

I already have.

What evidence did you ever provide that supported your claim that these numbers were supposedly fixed for their entire careers?

I have already posted this, too—the four sources from the FIA, wrc.com, Autosport and Speedcafe that all refer to "permanent numbers". I have posted those sources both here and in the WT:MOTOR discussion. They are also used in the 2019 World Rally Championship article to explain the regulation changes. I don't know how you keep missing these. Mclarenfan17 (talk) 22:06, 18 January 2020 (UTC)

And where exactly did my comment in that diff state "they have to re-apply at the end of each season". Sorry but these words are just not there. As I have explained multiple times, my point was that there was insufficient evidence that these numbers were fixed for there entire careers. And as multiple editors have pointed out to you time and time again during the WT:MOTOR discussion, your sources do NOT support your theory. None of them state that they are career numbers of numbers fixed for the entire careers. In fact the Autosport source literally uses the word season with regards to the numbers (Ogier's Citroen team-mate Esapekka Lappi will carry #4 in his first season with the French manufacturer...; A number was not assigned to Sebastien Loeb, who is currently competing on the Dakar Rally, at the unveiling, but the nine-time world champion's Hyundai will carry #19 on its six outings this season.). Yet for some reason you utterly refuse to accept that even though no one agrees with you.Tvx1 23:12, 18 January 2020 (UTC)

The part where you said:

It seems like they only reserve a number for the duration of a season.

So what happens at the end of that season if the number is only reserved for a year?

It's obvious what happened here. You never had the sources to support your claim, obviously engaged in original research, and now that you think you have a source that justifies it, you're trying to talk your way out of it. Mclarenfan17 (talk) 23:21, 18 January 2020 (UTC)

No, that's what you think I did. As explained multiple times. The only claim I actually made is that we had insufficient evidence to support the content. I you read incorrect things between the lines that's your problem, not mine. I can not be held accountable for your misinterpretations. If anyone engaged in original research and synthesis it was you. You included content in an article bases on a personal assumption causing to include incorrect information for weeks.Tvx1 15:50, 19 January 2020 (UTC)

Mclarenfan17, while you insisted us to provide sources that car numbers can change, you yourself failed to provide evidence that they stay the same. The sources you provided do not explain what is a "permanent number", and now with the number changes we know it was premature to write those wrong numbers in the first place. Content of Wikipedia should not be poorly sourced. You did the same thing with writing the "WRC-2 Pro" championship will run in 2020. Remember that you wrote in May that the championship will run, while is reality it does not. The false info stayed in the article for 6 months, when on 5th of October I finally removed it. Then you demanded sources and consensus from me. Talk:2020_World_Rally_Championship/Archive_1#WRC-2_Pro_in_2020 [4] [5]. Now I ask do you understand what All content must be verifiable. The burden to demonstrate verifiability lies with the editor who adds or restores material, and it is satisfied by providing an inline citation to a reliable source that directly supports[2] the contribution.[3] Any material lacking a reliable source directly supporting it may be removed and should not be restored without an inline citation to a reliable source means? Do you plan to write poorly sourced content in the future? Pelmeen10 (talk) 01:31, 19 January 2020 (UTC)

The sources you provided do not explain what is a "permanent number", and now with the number changes we know it was premature to write those wrong numbers in the first place.

The definition of permanent is "lasting or intended to last or remain unchanged indefinitely". Therefore, it is reasonable to assume that a "permanent number" is "a number that is lasting or intended to last or remain unchanged indefinitely".

Now I ask do you understand what All content must be verifiable

Do you? In the discussion that you cite, the only source you provided was an image of a Twitter feed. It was not clear who the author was, and totally failed WP:RELIABLE and WP:VERIFIABLE.

Do you plan to write poorly sourced content in the future?

I gave you four reliable, verifiable sources, which you ignored. And now you're trying to claim that "because those sources do not explicitly define what that term meant, I was under no obligation to provide any sources of my own".

You're trying to argue that because the word "permanent" was never defined in the sources, then we could not be sure that the word "permanent" actually meant "permanent" all to avoid the fact that you carried out original resesrch. Do I need a source that says the sky is blue as well? Mclarenfan17 (talk) 04:51, 19 January 2020 (UTC)

If you ever wanted a better example of someone trying to talk "their way out of it", this one from you here is the best yet.Tvx1 15:53, 19 January 2020 (UTC)