Talk:Gaza Health Ministry
This is the talk page for discussing improvements to the Gaza Health Ministry article. This is not a forum for general discussion of the article's subject. |
Article policies
|
Find medical sources: Source guidelines · PubMed · Cochrane · DOAJ · Gale · OpenMD · ScienceDirect · Springer · Trip · Wiley · TWL |
Archives: Index, 1Auto-archiving period: 90 days |
Warning: active arbitration remedies The contentious topics procedure applies to this article. This article is related to the Arab–Israeli conflict, which is a contentious topic. Furthermore, the following rules apply when editing this article:
Editors who repeatedly or seriously fail to adhere to the purpose of Wikipedia, any expected standards of behaviour, or any normal editorial process may be blocked or restricted by an administrator. Editors are advised to familiarise themselves with the contentious topics procedures before editing this page.
|
This article is rated C-class on Wikipedia's content assessment scale. It is of interest to multiple WikiProjects. | |||||||||||||||||||||||||||||||||||||||||
|
This article has been mentioned by a media organization:
|
Index
|
|
This page has archives. Sections older than 90 days may be automatically archived by ClueBot III when more than 5 sections are present. |
Extended-confirmed-protected edit request on 15 May 2024
editThis edit request has been answered. Set the |answered= or |ans= parameter to no to reactivate your request. |
Change sub-heading: Scientific studies
to
Casualty figure accuracy
Following the below paragraphs:
"The London School of Hygiene and Tropical Medicine conducted several analyses on the data and concluded it was "implausible" that GHM engaged in data fabrication.[21] The study found that GHM's reported crude mortality rate in the age bracket of 20-59 years was broadly similar to the mortality rate of UNRWA employees and the mortality rate of Gaza's health-care workers (reported by the World Health Organization).[21] The study also found that the number of buildings reported damaged by the Hamas-run Gaza Ministry of Public Works was consistent with satellite imagery-based estimates conducted by Sky News (both arrived at the figure of 7%).[21] The study looked at 7,028 reported deaths (Oct 7 to Oct 26), and found only one case of a duplicated identification number and one case of implausible age.[21]
Pennsylvania professor Abraham Wyner wrote in Tablet that the GHM casualty figures were "faked";[22] to which Columbia professor Les Roberts responded that GHM numbers were accurate and probably even an underestimate.[19] Wyner's main argument was that from Oct. 26 - Nov. 10, the number of deaths per day is 270 with "strikingly little variation".[22] CalTech statistician Lior Pachter responded that Wyner had cherrypicked a particular period, outside of which the variance was higher; even within Wyner's picked window the daily deaths had a standard deviation of 42.25 and variance of 1,785.[23] Wyner also argued that data showed lack of temporal correlation between total deaths, and those of women and children. In response, Marine Corps professor James Joyner quoted an opinion that GHM updates total deaths immediately, but there is a lag in updating the proportion of women and children, making time correlations "meaningless".[24]"
Add:
"An analysis of the data by the Telegraph found that the UNRWA figures showed a higher male casualty rate despite UNRWA consisting of primarily female staff, whereas the GHM casualty breakdown showed females constituting a higher casualty rate in the general population. They further noted that a third of the total UNRWA casualties occurred in the 7 days since October 7th whereas the casualty rate in the general population remained relatively consistent from October to January. The location of the deaths of UNRWA personnel was found to be largely equal between the North and the South of Gaza despite the IDF's focus being towards the North of Gaza. [1]
The Washington Institute for Near East Policy found that there were statistical inconsistencies in the GHM, OCHA and Gaza Media Office (GMO) fatality reports. There were 4 instances where the GHM reported adult male fatality figure was lower than the previous figure(s), and 3 instances where the reported adult female figure was lower than the previous figure(s). [2]
On the 8th May, the GHM and GMO reduced the reported adult female fatality figure from 9,500 to 4,959 and the children's fatality figure from 14,500 to 7,797 with the previously unreported adult male figure being reported to be 10,006.[3][4]" Aaronathers (talk) 00:32, 16 May 2024 (UTC)
- The stuff from Abraham Wyner is here because he has some credentials even though it is not quoted in reliable sources because it is trash statistics used for propaganda. The rest is nitpicking by people with no qualification and in unreliable sources, for instance Israel cut the number of civilians killed in the October 7 attack down by a few hundred, that is called checking the data and sometimes the Gaza health ministry makes mistakes like duplicating a batch once. We don't have bits saying that Israel cook their statistics or resurrect the dead because of that. NadVolum (talk) 09:49, 25 May 2024 (UTC)
- And by the way the death tolls were not reduced, what happened is hat the UN OCHR started reporting the number of fully identified casualties in the different categories and only giving the total figure for reported deaths. And even reported deaths is an underestimate of the total deaths because of those missing under the rubble. NadVolum (talk) 10:02, 25 May 2024 (UTC)
References
- ^ Lovett, Samuel; Corfe, Ollie; Nuki, Paul (4 February 2024). "Majority of UN workers killed in Gaza died while 'off duty,' data reveals". The Telegraph.
- ^ "How Hamas Manipulates Gaza Fatality Numbers: Examining the Male Undercount and Other Problems | The Washington Institute". www.washingtoninstitute.org. Retrieved 16 May 2024.
- ^ "United Nations Office for the Coordination of Humanitarian Affairs - occupied Palestinian territory | Hostilities in the Gaza Strip and Israel - reported impact | Day 213". United Nations Office for the Coordination of Humanitarian Affairs - occupied Palestinian territory. 6 May 2024. Retrieved 16 May 2024.
- ^ "United Nations Office for the Coordination of Humanitarian Affairs - occupied Palestinian territory | Hostilities in the Gaza Strip and Israel - reported impact | Day 215". United Nations Office for the Coordination of Humanitarian Affairs - occupied Palestinian territory. 8 May 2024. Retrieved 16 May 2024.
- Not done: per above. thetechie@enwiki: ~/talk/ $ 03:10, 2 June 2024 (UTC)
Wyner summary
edit@NadVolum: I strongly disagree with this removal. Without that, we're really misrepresenting the Wyner piece.
Wyner's arguments essentially fall into two categories, first that the data appeared too regular in some ways, and second that it appeared too irregular in other ways. After your change, we're mentioning the former (calling it the "main argument"), but not the latter. The latter frankly contains stronger arguments, and is the main focus on the piece.
You mention a counterargument about "batching", which seems like a rather unsubstantiated idea from a blog comment. Even if there was a more substantiated counterargument from a reliable source, it doesn't seem like a reason to remove the Wyner argument that it attempts to refute. — xDanielx T/C\R 17:14, 28 May 2024 (UTC)
- His main argument was that the increase was too regular and so must be made up - that affects the overall casualty figures which is what people are interested in. If his argument on that is patently false why bother with any of the rest of what he says since any professional statistician wouldn't make a mistake like that. There's no need for gish gallop. Do you actually believe Professor Wyner wrote some actual good statistics rather than writing a propaganda piece for Israel that any beginner in statistics can see through? It is a personal opinion piece which has not been peer reviewed and reliable sources didn't follow up on it. People can follow the link and read the whole sorry business up and be convinced by it if they are that way inclined. NadVolum (talk) 20:48, 28 May 2024 (UTC)
- Including two short sentences about Wyner's arguments in order to provide a reasonable (still very abbreviated) summary is very far from a "gish gallop".
- It would be a major NPOV violation to misrepresent the piece, cherry picking a single point for which a reasonable refutation exists, in order to make it appear as if the entire piece has been refuted.
- It doesn't matter whether you or I personally judge Wyner's arguments to be "convincing", "patently false", or whatever else. It's not our role as editors to make a determination about the merits of the argument.
- Why do you bring up peer review? It's not a requirement even for primary sources (there are plenty of options for secondary sources here too), and certainly not a justification for misrepresenting a source. — xDanielx T/C\R 03:31, 29 May 2024 (UTC)
- As to the second point it would have been a good idea for him to ask them why the anticorrelation happened, why didn't he? The batching by the type whether women men children or elderly as they checked them seems a very reasonable idea to me. NadVolum (talk) 21:12, 28 May 2024 (UTC)
- So you accept his main argument has problems. And he is a professional statistician who should check what he says before casting aspertions. And yet for some reason you think his second argument is better and you wish to put it in? Do you think you are a better qualified than him? Putting in a number of bad arguments in the hope that the quantity bamboozles people is the essence of Gish gallop and as noted there one should deal with the first argument first. The business about batching deals perfectly well with his second argument and there is no point going on to his third argument about bodies being resurrected. And then his fourth argument about the reported male deaths being suspiciously low. The title of that section says scientific. Only one real contradiction is needed to upset a scientific theory. There is no need for us to go through the lot. The piece has not had any peer review and is an obvious propaganda piece. we don't give a large amount of space to non-peer reviewed work saying peer reviewed work has got it wrong. NadVolum (talk) 13:47, 29 May 2024 (UTC)
- I don't have an opinion regarding the removal in question, but I wouldn't say that it's an "obvious propaganda piece". Several RS have mentioned it [1] [2]. Alaexis¿question? 19:56, 29 May 2024 (UTC)
- It seems rather uncharitable to ignore stronger points in the piece because a single point was questionable. In this case the other points didn't depend on the questionable one, so there's no broader theory that was upset.
- But in any case, I don't think it's our role to scrutinize the content of an essay which clearly merits inclusion based on relevance, significance and the author's credentials. — xDanielx T/C\R 04:05, 30 May 2024 (UTC)
- It simply isn't WP:DUE as it has very little weight. Editors don't have to into detail about every bit of rubbish that comes out. He gets paid in court cases as an expert witness on statistics, those egregious errors in his analysis are deliberate. NadVolum (talk) 12:35, 30 May 2024 (UTC)
- This material shouldn't appear under the subhead of "scientific studies" at all - an academic bypassing the peer review process by writing a methodologically suspect opinion piece in a highly partisan rag is not a scientific study; it is just about the opposite of one. Iskandar323 (talk) 13:42, 30 May 2024 (UTC)
- It's perfectly normal for analyses of recent events to not seek publication in peer-reviewed journals, a process which can take a year or more depending on the field and the preferred journals. Many respected economists self-publish analysis of recent events, for example.
- Agreed "scientific study" probably isn't the right label for it, but commentary doesn't seem quite fitting in my opinion for a somewhat technical analysis by a statistician. Let's just call it "analysis", it's how most sources refer to it. — xDanielx T/C\R 16:51, 30 May 2024 (UTC)
- Hatchet job? It is fairly elementary thankfully, he wasn't trying to fool experts. NadVolum (talk) 20:21, 31 May 2024 (UTC)
Wyner
edit@XDanielx: Can you elaborate on this edit? [3] I don't think some Wyner writing in a random magazine is due for any controversial claim. Also a weasel word has been introduced, and in WP voice, "apparent". Makeandtoss (talk) 14:44, 2 June 2024 (UTC)
- Hm I don't quite follow what you mean about due for any controversial claim?
- I agree the article'S inclusion In Tablet doesn't add much to its credibility, though At worst that means we should treat it like an WP:EXPERTSPS, which "may Be considered reliable when produced by an established subject-matter expert".
- My reasoning for the edit was mainly that it doesn't seem accurate to say the two Lancet articles refute Wyner'S article. They both predate Wyner'S article, so of course they don't mention Wyner, and they don't really look At the same things.
- In particular, Wyner'S article was largely about temporal trends. The LSHTM article doesn't look At temporal data, but evaluates other aspects of the data. The JHU article looks At temporal data specifically for GHM vs UNRWA, but not At other correlations such As men vs women. Even if there was a contradiction among the articles, we'd need an RS to establish that without SYNTH.
- To Be fair, Pachter'S blog does refute one of Wyner'S arguments (or At least aims to), but that'S all we seem to have; No RSs seem to engage Wyner'S separate points about irregularities.
- So I don't think we can frame Wyner'S arguments about irregularities As something that'S been refuted, though some kind of mild MOS:DOUBT language may Be appropriate, like "apparent" or "purported"? — xDanielx T/C\R 20:00, 2 June 2024(UTC)
- @XDanielx: This would Be WP:Fringe As this is a minority position In a magazine, refuted by findings of scientific journals and most of the concerned international organizations. Makeandtoss (talk) 09:53, 3 June 2024(UTC)
- I've split the sentence into two to avoid the improper synthesis. The irregularities found by Spagat aren't disproved and do not disprove the conclusions of the Lancet articles. Both are RS so we simply report what they say. Alaexis¿question? 12:39, 3 June 2024(UTC)
- Not In the lede As that would give undue weight. Makeandtoss (talk) 13:20, 3 June 2024(UTC)
- Please see my points above, e.g. on why the Lancet articles were not refuting Wyner'S. No RS has disputed Wyner'S points about irregularities. — xDanielx T/C\R 15:05, 3 June 2024(UTC)
- I removed the WP:UNDUE bit In the lead. It was stated In Wiki voice and pitted against a peer reviewed source. NadVolum (talk) 17:15, 3 June 2024(UTC)
- They'Re not arguing against one another though, they make largely unrelated claims which can both Be true, As I discussed above.
- It'S like if one article said that New York has nice food and culture, while another said that the weather is poor. In some sense they'Re supporting different narratives, but they'Re not contradicting one another, and the credibility of one article wouldn't imply anything about the other.
- It seems like a major NPOV issue to devote the entire second paragraph to the narrative that the data is reliable, while removing the one and only mention of concerns about the quality of the data. — xDanielx T/C\R 17:51, 3 June 2024(UTC)
- The point is that it is not peer reviewed and it is saying there are problems when a peer reviewed one has said it is okay. It does not matter that you have this idea that it is reliable, it is simply WP:UNDUE given WP:SCHOLARSHIP. The article has about his saying there are problems with the correlation between men and women but He simply should not be in the lead at all. That's like saying in the lead about a drug that a study showed no problems but somebody wrote there might be some problems with it. Especiallly when that somebody didn't do any checking with the people who wrote the study or anyone independent. That is pretty standard on X or TikTok but it is not what Wikipedia does. NadVolum (talk) 19:34, 3 June 2024 (UTC)
- This is not a policy-based argument. Michael Spagat is a well-known researcher of wartime casualties and his work is reliable and notable even if it's not published in a scholarly journal. Alaexis¿question? 20:22, 3 June 2024 (UTC)
- The lead was referring to Wyner's article which is trash. But yes Michael Spagat's stuff should also not be in the lead till it is checked by somebody. Please read WP:SCHOLARSHIP about the relative merits of works that are peer reviewed compared to those that aren't. NadVolum (talk) 20:32, 3 June 2024 (UTC)
- There is no rule that the lede should contain only information found in peer-reviewed studies (it doesn't). There is a rule that the lede should summarise the article. The article mentions several alleged inconsistencies, and we should briefly summarise them in the lede. Alaexis¿question? 21:22, 3 June 2024 (UTC)
- I am not part of this discussion, but correction please. The article only discusses one allegation of falsified casualty numbers via a Tablet article by Abraham Wyner, rather than several alleged inconsistencies – hence why I clarified that in the lede earlier. It should also be further clarified that the Tablet article suggests that the data is largely inaccurate/fake, not just there being "some irregularities observed," as there is a large difference between stating that there are some irregularities in a piece of data, and suggesting that it is either largely inaccurate, outright fake, or both. B3251 (talk) 01:09, 4 June 2024 (UTC)
- The Tablet stuff is straight prpaganda with twisted statistics by professor Wyner. Professor Michael Spagat wrote https://aoav.org.uk/2024/analysis-of-new-death-data-from-gazas-health-ministry-reveals-several-concerns/ which points out some problems but he seems honest and is an expert on this sort of matter. We shouldn't be putting what they wrote into the lead alongside peer reviewed work. It can be checked fairly quickly by somebody like the peer reviewed stuff was and can wait till then. That it is later does not mean it overrides peer reviewed work. NadVolum (talk) 07:12, 4 June 2024 (UTC)
- You're entitled to your opinion about Wyner, though these don't seem like an objective concerns backed by reliable sources.
- Neither overrides the other since they look at completely separate aspects of the data. The Lancet studies are also rather old now; only one looked at the time period that Wyner's concerns were about, and neither looked at the time period Spagat's concerns were about. They're still relevant, but in no way do they contradict Wyner's or Spagat's concerns. — xDanielx T/C\R 18:04, 4 June 2024 (UTC)
- That's your opinion. The OCHA says it is happy about the figures and as I said before those other sources are undue in the lead when there are peer reviewed sources. If you want to start an RfC then do so, otherwise we should follow policy about this. NadVolum (talk) 18:13, 4 June 2024 (UTC)
- I also don't agree with putting any of the Wyner, Pachter, Roberts or Spagat stuff in the lead. I find the lead, at the time of writing this comment, a sufficient summary of the body. — hako9 (talk) 18:33, 4 June 2024 (UTC)
- The Tablet stuff is straight prpaganda with twisted statistics by professor Wyner. Professor Michael Spagat wrote https://aoav.org.uk/2024/analysis-of-new-death-data-from-gazas-health-ministry-reveals-several-concerns/ which points out some problems but he seems honest and is an expert on this sort of matter. We shouldn't be putting what they wrote into the lead alongside peer reviewed work. It can be checked fairly quickly by somebody like the peer reviewed stuff was and can wait till then. That it is later does not mean it overrides peer reviewed work. NadVolum (talk) 07:12, 4 June 2024 (UTC)
- I am not part of this discussion, but correction please. The article only discusses one allegation of falsified casualty numbers via a Tablet article by Abraham Wyner, rather than several alleged inconsistencies – hence why I clarified that in the lede earlier. It should also be further clarified that the Tablet article suggests that the data is largely inaccurate/fake, not just there being "some irregularities observed," as there is a large difference between stating that there are some irregularities in a piece of data, and suggesting that it is either largely inaccurate, outright fake, or both. B3251 (talk) 01:09, 4 June 2024 (UTC)
- There is no rule that the lede should contain only information found in peer-reviewed studies (it doesn't). There is a rule that the lede should summarise the article. The article mentions several alleged inconsistencies, and we should briefly summarise them in the lede. Alaexis¿question? 21:22, 3 June 2024 (UTC)
- The lead was referring to Wyner's article which is trash. But yes Michael Spagat's stuff should also not be in the lead till it is checked by somebody. Please read WP:SCHOLARSHIP about the relative merits of works that are peer reviewed compared to those that aren't. NadVolum (talk) 20:32, 3 June 2024 (UTC)
- This is not a policy-based argument. Michael Spagat is a well-known researcher of wartime casualties and his work is reliable and notable even if it's not published in a scholarly journal. Alaexis¿question? 20:22, 3 June 2024 (UTC)
- The point is that it is not peer reviewed and it is saying there are problems when a peer reviewed one has said it is okay. It does not matter that you have this idea that it is reliable, it is simply WP:UNDUE given WP:SCHOLARSHIP. The article has about his saying there are problems with the correlation between men and women but He simply should not be in the lead at all. That's like saying in the lead about a drug that a study showed no problems but somebody wrote there might be some problems with it. Especiallly when that somebody didn't do any checking with the people who wrote the study or anyone independent. That is pretty standard on X or TikTok but it is not what Wikipedia does. NadVolum (talk) 19:34, 3 June 2024 (UTC)
- I removed the WP:UNDUE bit In the lead. It was stated In Wiki voice and pitted against a peer reviewed source. NadVolum (talk) 17:15, 3 June 2024(UTC)
- I've split the sentence into two to avoid the improper synthesis. The irregularities found by Spagat aren't disproved and do not disprove the conclusions of the Lancet articles. Both are RS so we simply report what they say. Alaexis¿question? 12:39, 3 June 2024(UTC)
- @XDanielx: This would Be WP:Fringe As this is a minority position In a magazine, refuted by findings of scientific journals and most of the concerned international organizations. Makeandtoss (talk) 09:53, 3 June 2024(UTC)
- @XDanielx: I'm very disappointed by this edit. You want to include Wyner but want to suppress criticism of Wyner? Either we include both or remove this altogether (I prefer including).VR (Please ping on reply) 22:00, 4 June 2024 (UTC)
- Surely that can't be considered a reliable source - it's a comment on a blog written by "Ken M". Besides that it just seems like unsubstantiated speculation. There's no quote from the GHM or anyone with knowledge of GHM's processes, just an assumption by "Ken M" about what those processes might look like.
- I'm normally all for covering all sides of an argument, but just don't see how this one could be included without a blatant WP:V and WP:RS violation. Hopefully a better source will turn up. — xDanielx T/C\R 17:12, 5 June 2024 (UTC)
- Wyner is in practically the same boat. They are both professors. Abraham Wyner has no particular expertise in war or casualties, James Joyner is not an expert in statistics. What they have published is just their own opinions. Either both statements should go or both stay. You may think it is unsubstantiated speculation - but Wyner should have checked and he didn't. He just invented spurious arguments. NadVolum (talk) 18:30, 5 June 2024 (UTC)
- James Joyner might be a reliable source, but he merely compiled a list of quotes from anonymous internet users, without really endorsing or even commenting on them. The argument isn't coming from him at all, so whatever credibility he might have seems irrelevant. — xDanielx T/C\R 00:21, 6 June 2024 (UTC)
- You said 'It's not our role as editors to make a determination about the merits of the argument.' And yet you think this is more credible than Wyner's main argument and so want to put it in. As to Joyner he assessed the argument and said "Additionally, commenter Ken M adds this insight", that isn't just compiling a list, it is an assessment of the argument. NadVolum (talk) 13:51, 6 June 2024 (UTC)
- A few points here,
- I wouldn't call Wyner's point about minimal variation his "main argument". It's only two paragraphs, framed as "the first place to look". Wyner's article follows a sort of narrative arc, starting with slightly suspicious things and building up to more clearly problematic things, like the men vs women anticorrelation.
- Pachter only "refutes" the visual anyway, not the introductory argument itself. Pachter's concern was solely about looking at cumulative data. The actual text of Wyner's argument talks about "variation day to day" or "the daily reported casualty count", not about its cumulative sum.
- Where did Joyner "assess" the argument by Ken M? I guess using the word "insight" counts for something, but a one-word label with no further discussion isn't much of an assessment.
- — xDanielx T/C\R 17:01, 6 June 2024 (UTC)
- This is the twenty first century, not the eighteenth where they liked long arguments as one of their entertainments. This is the century of the gnatlike attention span and flick left or right. He has jobs as an expert witness and knows that. I think citation needed for the 'story-arc' idea. And yes I think saying insght and quoting something does count. Far more than the assessment of someone on Wikipedia. And his finishing comments are not exactly supportive of Wyner's article either. NadVolum (talk) 18:24, 6 June 2024 (UTC)
- But you removed the counterpoint for the (alleged) men vs women anti-correlation too. If that's his main point, why shield him from criticism on it? And Joyner quoting an opinion does give us the necessary sourcing for inclusion.VR (Please ping on reply) 22:58, 13 June 2024 (UTC)
- A few points here,
- You said 'It's not our role as editors to make a determination about the merits of the argument.' And yet you think this is more credible than Wyner's main argument and so want to put it in. As to Joyner he assessed the argument and said "Additionally, commenter Ken M adds this insight", that isn't just compiling a list, it is an assessment of the argument. NadVolum (talk) 13:51, 6 June 2024 (UTC)
- James Joyner might be a reliable source, but he merely compiled a list of quotes from anonymous internet users, without really endorsing or even commenting on them. The argument isn't coming from him at all, so whatever credibility he might have seems irrelevant. — xDanielx T/C\R 00:21, 6 June 2024 (UTC)
- James Joyner is Professor and Department Head of Security Studies at Marine Corps University's Command and Staff College. He's a former Army officer and Desert Storm veteran. I think he is reliable enough to respond to Wyner. VR (Please ping on reply) 16:10, 9 June 2024 (UTC)
- Wyner is in practically the same boat. They are both professors. Abraham Wyner has no particular expertise in war or casualties, James Joyner is not an expert in statistics. What they have published is just their own opinions. Either both statements should go or both stay. You may think it is unsubstantiated speculation - but Wyner should have checked and he didn't. He just invented spurious arguments. NadVolum (talk) 18:30, 5 June 2024 (UTC)
- I think Wyner is given way too much weight for a paper in a partisan barely reliable source, and his past statements about global warming being debunked, along with his avowed partisanship in this topic, make him considerably less useful. When we have peer-reviewed scholarship we dont "balance" it with partisan talking points that have not been reviewed by anybody. nableezy - 18:27, 6 June 2024 (UTC)
New AP article with good info that could help the article
edithttps://apnews.com/article/israel-palestinians-hamas-war-casualties-toll-65e18f3362674245356c539e4bc0b67a 57.140.28.26 (talk) 22:35, 7 June 2024 (UTC)
- Two interesting things I observed. The analysis says "Only unique, fully identified deaths are counted." It admits that it has not included 29% of deaths because some bodies are "not claimed by families, decomposed beyond recognition or whose records were lost in Israeli raids on hospitals." The second is that, for the first time, we are given Hamas-reported figure of how many militants have been killed. It estimates the number to be 6,000.VR (Please ping on reply) 16:23, 9 June 2024 (UTC)
- There's a bit of a problem with that in that the hospitals used to identify all the deaths but now many are done using forms filled in by the pubic since only a fraction of the hospitals still do that. A problem with that is that widows need to fill in details of their husbands to get state assistance but there is far less incentive to identify other deaths. The business about drones and the IDF killing instea dof using bombs probably is true to an extent - but exactly to what extent compared to it being an artifact of the reporting is anybody's guess. On the other hand in the past possibly fewer men were identified as their bodies would be less likely to be accessible. NadVolum (talk) 16:52, 9 June 2024 (UTC)
- The 6,000 figure has been around for a while but has nothing to do with the Gaza Health Ministry. See Casualties of the Israel–Hamas war#Civilian to military ratio. In fact probably all that AP article has more to do with that other Wikipedia article. NadVolum (talk) 16:59, 9 June 2024 (UTC)
- @NadVolum: yes, but the date matters. Here Hamas is giving the figure at a time when the toll is much higher. The earlier 6,000, given months ago, was officially denied by the group.VR (Please ping on reply) 23:04, 13 June 2024 (UTC)
- Well I'll try adding it to the casualties article in the section about militants and see how that goes. NadVolum (talk) 18:47, 14 June 2024 (UTC)
- @NadVolum: yes, but the date matters. Here Hamas is giving the figure at a time when the toll is much higher. The earlier 6,000, given months ago, was officially denied by the group.VR (Please ping on reply) 23:04, 13 June 2024 (UTC)
The Lancet articles not peer reviewed?
editThe two The Lancet articles being peer reviewed has been mentioned 17 times here, and frequently cited as a reason for elevating them above sources that question GHM's reliability.
They don't appear to be peer reviewed though, and "scientific studies" might be a somewhat misleading label. The Telegraph says of one article,
Because the work was published as correspondence rather than a formal study, it was not peer reviewed [...]
The other article is also under "correspondence". The Lancet describes "correspondence" as (emphasis theirs)
Our readers’ reflections on content published in the Lancet journals or on other topics of general interest to our readers. These letters are not normally externally peer reviewed.
— xDanielx T/C\R 18:21, 9 June 2024 (UTC)
- Sorry didn't check. Lancet is still considered a stellar source on Wikipedia and it has a strong editorial policy. If you really want to put Wyner in the lead to go against those then go to WP:RSN or set up a WP:RfC here as it simply won't stick there otherwise. NadVolum (talk) 19:36, 9 June 2024 (UTC)
- I don't think non-biomedical claims and stories/analysis of news/current events are peer-reviewed very often. Lancet says their editorials are also not externally peer reviewed. So for instance, [4] isn't peer-reviewed. So is it worse than "The Tablet"? lol. — hako9 (talk) 20:47, 9 June 2024 (UTC)
- Maybe let's do a RfC: how should the lede summarise the "Scientific studies" and "Other analysis" sections with the options
- Only mention the Lancet studies (current version)
- Write that "some irregularities were observed in the data related to the Israel-Hamas war" ([5])
- Summarise it in a different way.
- Alaexis¿question? 11:38, 10 June 2024 (UTC)
- "Some irregularities" is weasel wording. Wyner has been debunked. But of course there have been other concerns like lack of identification numbers.VR (Please ping on reply) 23:00, 13 June 2024 (UTC)
One of the worst abuses of statistics I've ever seen
editJoshua Loftus, professor of statistics and data science at LSE, saying "One of the worst abuses of statistics I've ever seen" and "shockingly irresponsible" about the Wyner article on X at https://x.com/joftius/status/1766199967364890949 - is this okay for inclusion? NadVolum (talk) 10:08, 10 June 2024 (UTC)
- Seems to fail WP:BLPSPS, as WP:RSPTWITTER reminds us. The latter also says "Tweets that are not covered by reliable sources are likely to constitute undue weight"; I think this fails that as well although there's more room for interpretation there.
- Even if it wasn't strictly a policy violation, documenting Twitter or blog comment arguments just doesn't feel encyclopedic, and I don't understand the motivation to make the Wyner article appear thoroughly refuted at all costs. — xDanielx T/C\R 03:03, 11 June 2024 (UTC)
- Have you ever considered that Loftus and those other professors might be right and you're wrong? NadVolum (talk) 07:05, 11 June 2024 (UTC)
- @NadVolum: I would have been ok to Loftus if he had anything substantial to say, some mathematical counter-point.VR (Please ping on reply) 23:02, 13 June 2024 (UTC)
- @Vice regent: The others say why it is bad but really it is the opinion of other experts that counts on Wikipedia rather than their reasoning. What Loftus said gets to the bottom of it - the use and abuse of statistics is a common topic and his opinion is this is a dreadful example of the abuse of statistics. NadVolum (talk) 09:17, 14 June 2024 (UTC)
- Basically this is saying Wyner has gone way over to the dark side as far as principle D in https://www.amstat.org/your-career/ethical-guidelines-for-statistical-practice is concerned. NadVolum (talk) 09:25, 14 June 2024 (UTC)
- But it’s only a tweet and has received no secondary coverage. ꧁Zanahary꧂ 00:04, 7 July 2024 (UTC)
- If there is anything substantial, such as them publishing an article or even a blog/self-published source, about their statistical analysis that led them to the conclusion that it's "one of the worst abuses of statistics", then sure, their view can be included. We are not going to include someone's private musings in one tweet when they don't actually provide any substance to that opinion, unless that substance is provided in other reliable sources. We aren't a compendium of everything anyone remotely an expert has ever said on a subject - we do report on what experts say, but only when their opinion is either widely cited by others, or when their status as a significantly leading expert is such that their opinion on its own is notable (ex: we would consider covering the head of the CDC's opinion on COVID topics, even if their opinion was not itself covered significantly by others, but again, that's just a consideration). Overall, no reason to include based on one tweet, and if he expands on his opinion in another self-published source or journal article, then that would be what should be cited for inclusion, not the tweet. -bɜ:ʳkənhɪmez (User/say hi!) 01:14, 7 July 2024 (UTC)
Revisiting Wyner
editI hate to bring this up again, but it seems like the Wyner section has gotten even worse, creating serious WP:BLP, WP:RS and other issues -
- Roberts' article is framed as refutation of Wyner. While Roberts does mention "a recent article in Tablet" in passing, it doesn't engage the content of Wyner's article at all, so this is misleading.
- Wyner's argument about minimal variation is referred to as his "main argument". I wouldn't call it that and I don't see what basis there is for calling it that.
Pachter responded that Wyner had cherrypicked a particular period
is misleading, since this was an in-passing remark Patcher made in the comment section of his blog. Quoting such remarks from the comment section is probably WP:UNDUE and just feels unencyclopedic. At a minimum, this should be framed differently to emphasize that this is not the point of Pachter's post, which was a criticism of one of Wyner's visualizations.- Worse still, we quote another blog comment from an anonymous "Ken M". The argument for keeping this has been that the comment is quoted in another blog by James Joyner. However, Joyner doesn't discuss the argument; he simply copy/pasted it into a compilation of quotes. Even if the argument was coming from Joyner (which it isn't), Joyner doesn't seem like
subject-matter expert
as required by WP:BLOGS (the subject matter here isn't political science or the US army).
It seems like there may also be WP:BLPSPS issues here, although I'm not entirely clear on when criticism of a subject's work is considered information about the subject
(I asked a related question here). — xDanielx T/C\R 21:02, 7 July 2024 (UTC)
- I think I better just remove all mention of Wyner altogether as what he wrote was not an analysis and describing what he actually did would cause BLP problems. NadVolum (talk) 20:19, 9 July 2024 (UTC)
- What do you mean by not an analysis?
- There's already a lack of WP:BALANCE here, e.g. with the lede not mentioning any concerns about the data. Removing Wyner, the only real critical analysis of GHM data to appear in reliable sources, would make this much worse.
- I don't think there's a BLP violation as long as we don't include the blog comments. Wyner and Patcher do have secondary coverage, avoiding WP:BLPSPS. — xDanielx T/C\R 20:39, 9 July 2024 (UTC)
- As has been made clear by other academics in the same field and with a higher reputation that they think what he wrote is deliberate bad statistics. Do you want an RfC on whether Wyner's article should be considered as a statistical analysis or as an abuse of statistics? NadVolum (talk) 20:51, 9 July 2024 (UTC)
- I don't think insults among scholars on Twitter (or elsewhere) are a useful signal. Morris called Pappe's work "truly appalling". Finkelstein called Dershowitz's work "a collection of fraud, falsification, plagiarism and nonsense". At the end up the day they're still prominent scholars. — xDanielx T/C\R 21:07, 9 July 2024 (UTC)
- As has been made clear by other academics in the same field and with a higher reputation that they think what he wrote is deliberate bad statistics. Do you want an RfC on whether Wyner's article should be considered as a statistical analysis or as an abuse of statistics? NadVolum (talk) 20:51, 9 July 2024 (UTC)
- I take it you will continue on with this business unless an RfC says otherwise since you have ignored the answers at NPOVN. I will therefoe raise an RfC on the business. NadVolum (talk) 21:07, 9 July 2024 (UTC)
- From reviewing all the past discussions (not any particular WP:LOCALCONSENSUS), I believe the broad consensus has been roughly in the direction of inclusion (not prominently, but as a reasonable minority view).
- If you disagree, no objection to an RfC to clarify what the broader consensus is on this. — xDanielx T/C\R 21:11, 9 July 2024 (UTC)
- And I have removed it meantime. Selfstudier (talk) 22:02, 9 July 2024 (UTC)
- I take it you will continue on with this business unless an RfC says otherwise since you have ignored the answers at NPOVN. I will therefoe raise an RfC on the business. NadVolum (talk) 21:07, 9 July 2024 (UTC)
RfC - criticism of Abraham Wyner's article
editAbraham Wyner, a professor of statistics, wrote in The Tablet that the Gaza Health Ministry casualty figures were "faked". There has been a number of discussions disagreeing about what can be included. Four academics who have entries on Wikipedia including two of statistics have dismissed the article but only one has written an analysis and that of only the first argument. Should comments which don't include analysis be included? The latest discussions are at WP:NPOVN#Gaza Health Ministry, Talk:Gaza Health Ministry#Revisiting Wyner, Talk:Gaza Health Ministry#One of the worst abuses of statistics I've ever seen, Talk:Gaza Health Ministry#Wyner, Talk:Gaza Health Ministry#Wyner summary, Talk:Gaza Health Ministry/Archive 1#Extended-confirmed-protected edit request on 4 May 2024, WP:Reliable sources/Noticeboard/Archive 431#Tablet (magazine) and article by Wharton statistician. NadVolum (talk) 22:00, 9 July 2024 (UTC)
Discussion
edit- @NadVolum: the scope seems a bit unclear to me. Since this RfC arose from the deletion of the Wyner content altogether, I assume this RfC is meant to gauge consensus on whether Wyner content should be included at all, in addition to whether certain critiques of Wyner should be included? I think some more clarification would be helpful. In my view it's most important to answer whether Wyner should be included at all; possible inclusion of Tweets etc. then become more clear questions of policy which can be discussed on places like WP:RSN if needed. — xDanielx T/C\R 00:03, 10 July 2024 (UTC)
- If the Wyner source is included, his critics should have a right of reply. I'm neutral on whether to include him in the first place. His arguments do not appear particularly persuasive or significant to me. Frankly, he seems like a crank with a history of misrepresenting statistics to prove his preconceived point. However if enough other editors think it's important to cite him that's fine with me so long as the criticisms of his view are included for context. Unbandito (talk) 02:17, 10 July 2024 (UTC)
- He is a crank. See below, but he has used statistics to advocate against Covid boosters for children and against California ethnic studies requirements. We don't include his opinions as "statistical evidence" for either the COVID-19 vaccine nor would we include him as a source for a california ethnic studies class if it existed. I see no reason to put a single crank until either others corroborate his work. There's plenty of other ways to include opinions about whether the GHM is reliable or not. Bluethricecreamman (talk) 03:20, 10 July 2024 (UTC)
- Re
Four academics who have entries on Wikipedia including two of statistics"
, I'm not sure it's a fair summary. Here 3 academics say that "Several of these oddities have recently been noted ... by ... Abraham Wyner". Alaexis¿question? 22:02, 10 July 2024 (UTC)
Alternatives
editUser:Alaexis mentioned this article in Fathom, which makes some similar points as Wyner. Would those opposing the inclusion of Wyner be fine with including this as an alternative? Some notes -
- Fathom is published by BICOM. The editor is Alan Johnson.
- Fathom seems generally pro-Israel, but AFAIK nothing extreme. Fathom mentions a pro-2ss stance in their founding statement, mirroring Johnson's views.
- I haven't looked into the three authors much, but looks like they're PhDs in somewhat relevant fields.
- This article has fairly minor coverage, in somewhat niche sources such as JC. Wyner has more coverage.
- That said, content isn't required to be notable, and WP:WEIGHT is concerned with the signifiance of the overall viewpoint rather than the particular source we use to represent it. So if editors feel this source is more credible than Wyner, it could work as an alternative source for (broadly speaking) the same viewpoint.
- Some of the article's point involve GMO publications as well as MoH publications, so we'd need to be a bit careful about which bits of the article we mention.
— xDanielx T/C\R 21:02, 11 July 2024 (UTC)
- (NYT) Fighting Isn’t the Only Killer of Gazans Amid the War, Researchers Say Selfstudier (talk) 21:37, 11 July 2024 (UTC)
- I don't follow the connection? — xDanielx T/C\R 22:11, 11 July 2024 (UTC)
- Yet another source confirming the reliability of GHM, currently I am not disposed to accept Israeli propaganda to the contrary, regardless where it comes from. Selfstudier (talk) 22:47, 11 July 2024 (UTC)
- That recent Lancet essay didn't examine GHM reliability at all, beyond an introductory comment that
they have been accepted as accurate
. The rest is about indirect deaths, which doesn't have much relevance since it's not something GHM reports. - The two Lancet articles mentioned on this main page are more relevant, but still don't contradict Wyner or the Fathom article, since they're looking at entirely different aspects of the data. The word "women" doesn't even appear in one of them.
- NPOV requires us to
represent all significant viewpoints
. If your position is that you categorically oppose mentioning this significant minority viewpoint, regardless of which source we use and how reliable it is, that seems quite difficult to justify from a policy perspective. — xDanielx T/C\R 23:17, 11 July 2024 (UTC)- NYT also says it is reliable, which it is, there are numerous reliable sources confirming that GHM is reliable and Wyner saying it is not, so no contest. Selfstudier (talk) 09:16, 12 July 2024 (UTC)
- The 'alternative' does not seem to have anything like the coverage of Wyner's article and is in a place where any qualified people are unlikely to look. I think we should wait till someone actually evaluates it. What are we supposed to do - stick in the latest stuff like when Wyner's came out about the GHM, leave it in until like Wyner's article you get some actual feedback and then fight about the feedback until yet another alternative comes along? If they had got it into Lancet or something else with some reputation perhaps there would be some response quicker - like there has been already for instance for the Lancet one above about indirect deaths. Enough debates have already been spent on Wyner's article - it is time to finish debate on it one way or another. The 'alternative' can be dealt with on its own merits as a separate issue and not as if Wikipedia was a bargaining salon. NadVolum (talk) 10:37, 12 July 2024 (UTC)
- Yes, I can't believe there has been this much discussion over half-baked opinion piece in a low-grade online magazine. It's weight is zilch. It's credibility is shot. It's been ripped to shreds all over the web. And it has no significance. Iskandar323 (talk) 15:57, 12 July 2024 (UTC)
- Holding off on including it because it hasn't received much scrutiny seems like an unusual standard that isn't normally applied elsewhere. Normally if a source is advancing a controversial opinion, we just attribute it and maybe add some MOS:DOUBT language. — xDanielx T/C\R 17:12, 12 July 2024 (UTC)
- If that was what we normally did, Wyner would be included already. Selfstudier (talk) 17:17, 12 July 2024 (UTC)
- That recent Lancet essay didn't examine GHM reliability at all, beyond an introductory comment that
- Yet another source confirming the reliability of GHM, currently I am not disposed to accept Israeli propaganda to the contrary, regardless where it comes from. Selfstudier (talk) 22:47, 11 July 2024 (UTC)
- I don't follow the connection? — xDanielx T/C\R 22:11, 11 July 2024 (UTC)
- In what way is this relevant or an "alternative" to the Wyner article under discussion here? IOHANNVSVERVS (talk) 17:58, 12 July 2024 (UTC)
- It's a different source which makes some similar arguments, such as pointing out decreases in implied male deaths. It represents a broadly similar viewpoint, so if editors have concerns about the personal credibility of Wyner, this would be an alternative means of represent[ing] all significant viewpoints while avoiding any such concerns. — xDanielx T/C\R 18:15, 12 July 2024 (UTC)
- "Generally, the views of tiny minorities should not be included at all".
- We need to be searching for the best sources we can find and then presenting what they state. We should not be searching for the best sources we can find in support of a predetermined POV. IOHANNVSVERVS (talk) 18:21, 12 July 2024 (UTC)
- It's a significant minority in the niche area of GHM accuracy for the present conflict. The Fathom article could probably be considered one of just three scholarly works in this niche, depending on where we draw the line. — xDanielx T/C\R 19:02, 12 July 2024 (UTC)
- This whole 'alternatives' section should be closed as being off-topic. This RfC is about Wyner's article, not about the viewpoint that GHM's casualty reports are unreliable. IOHANNVSVERVS (talk) 23:13, 12 July 2024 (UTC)
- It's a significant minority in the niche area of GHM accuracy for the present conflict. The Fathom article could probably be considered one of just three scholarly works in this niche, depending on where we draw the line. — xDanielx T/C\R 19:02, 12 July 2024 (UTC)
- It's a different source which makes some similar arguments, such as pointing out decreases in implied male deaths. It represents a broadly similar viewpoint, so if editors have concerns about the personal credibility of Wyner, this would be an alternative means of represent[ing] all significant viewpoints while avoiding any such concerns. — xDanielx T/C\R 18:15, 12 July 2024 (UTC)
- The problem with the Fathom article is that it's entire analysis is based on the premise that GHM claims the 70%, and it is meant as a precise number. This is a false assumption. RS have said the 70% comes from GMO not GHM[6]. Consider this statement by Michael Spagat: "The GMO regularly releases numbers that assign roughly 70% of all deaths to women and children. Yet the corresponding MoH percentages, currently 54.4%, are much lower." It shows us that the 70% estimate was always meant to be "rough" and made by the GMO not GHM.VR (Please ping on reply) 17:05, 25 July 2024 (UTC)
!Votes
edit- My !vote is to include a mention of the Wyner article, as it's a significant minority view. There's nothing fringe about Wyner's article, which raised reasonable concerns about GHM data, like an unexplained strong negative correlation between the daily casualties of men and women. Despite efforts to frame the article as debunked (see the previous section of this talk page), Wyner's questions remain largely unanswered. Pachter criticized one of the visualizations as misleading, but did not refute the rest of the article. Pachter's post itself should be mentioned, but not arguments in the comment section. Two Lancet essays (not subject to peer review) did not identify significant issues with GHM data, but they were examining entirely different aspects of the data, as well as different time periods. It remains the majority view that GHM data is generally reliable, so I'm not arguing for equal weight, just some inclusion of this significant minority view in the interest of balance. — xDanielx T/C\R 22:38, 9 July 2024 (UTC)
- Largely unanswered only in the sense that the answers are not in a form that can be used on Wikipedia. A responder to Pachter's blog for instance gave a perfectly good possibility in that the checking of deaths was done in separate batches for men, women and children and that would completely explain the strong negative correlation. This should have been an obvious possibility for Wyner but he didn't check and said 'This lack of correlation is the second circumstantial piece of evidence suggesting the numbers are not real'. What statistician finds possible problems and just writes it up that way without checking? And do you really think the GMH would put out figures like that if they were really making them up and according to Wyner had the overall figures going up too smoothly? That's just daft. NadVolum (talk) 08:58, 10 July 2024 (UTC)
- Re the "Ken M" theory, the GHM hasn't mentioned anything about a delayed process like that. It seems reasonable for Wyner to take the published data at face value, and not speculate about hypothetical GHM processes that might explain errors in the published dates. Perhaps he could have asked GHM for an explanation, but he might have no means of contacting them. — xDanielx T/C\R 21:44, 10 July 2024 (UTC)
- GHM hasn't mentioned that, but a Professor and Department Head of Security Studies at Marine Corps University's Command and Staff College has.VR (Please ping on reply) 18:25, 13 July 2024 (UTC)
- His background isn't that relevant, and he didn't corroborate the "Ken M" theory, he just copy/pasted it as part of a compilation of comments he found interesting. — xDanielx T/C\R 20:33, 13 July 2024 (UTC)
- There is no requirement that a source add any extra analysis. In law, for example, jurists quote certain opinions all the time without further commentary.VR (Please ping on reply) 19:11, 16 July 2024 (UTC)
- His background isn't that relevant, and he didn't corroborate the "Ken M" theory, he just copy/pasted it as part of a compilation of comments he found interesting. — xDanielx T/C\R 20:33, 13 July 2024 (UTC)
- "At face value" here appears to be a euphemism for "without considering any potential variables" and "in complete ignorance of any of the facts on the ground" – no, that is in fact not reasonable; quite the opposite. Iskandar323 (talk) 20:39, 13 July 2024 (UTC)
- That seems like a peculiar way to describe speculation about hypothetical processes. Even if these hypothetical processes did exist, at best they might explain flaws in the data. It seems reasonable to simply point out such flaws without speculating about internal GHM processes (which only they are privy to) that might explain the flaws. — xDanielx T/C\R 15:08, 14 July 2024 (UTC)
- What are these flaws that you speak of? Are these the errors that fall into the category of expected human error in a health sector under siege, or are you taking about peculiar artefacts of the data? If it is the latter, the notion that these are "flaws" in the data at all is an assumption and speculation – one that is based on the presumption that these artefacts in the data are not readily explained by the ways in which the data is gathered and recorded. Iskandar323 (talk) 15:51, 14 July 2024 (UTC)
- The "Ken M" theory seems to be that after GHM goes through some (hypothetical) delayed process for determining age and gender, they somehow forget the date of the recorded death, or mix it up with the date that the process concluded.
- It's certainly understandable if GHM can't collect precise data given the circumstances, though their use of exact numbers (cf. significant digits) and dates creates an expectation of relative precision. — xDanielx T/C\R 05:13, 15 July 2024 (UTC)
- Let's stay on the topic of the Wyner article specifically @Iskandar323 and @XDanielx. IOHANNVSVERVS (talk) 05:24, 15 July 2024 (UTC)
- What are these flaws that you speak of? Are these the errors that fall into the category of expected human error in a health sector under siege, or are you taking about peculiar artefacts of the data? If it is the latter, the notion that these are "flaws" in the data at all is an assumption and speculation – one that is based on the presumption that these artefacts in the data are not readily explained by the ways in which the data is gathered and recorded. Iskandar323 (talk) 15:51, 14 July 2024 (UTC)
- That seems like a peculiar way to describe speculation about hypothetical processes. Even if these hypothetical processes did exist, at best they might explain flaws in the data. It seems reasonable to simply point out such flaws without speculating about internal GHM processes (which only they are privy to) that might explain the flaws. — xDanielx T/C\R 15:08, 14 July 2024 (UTC)
- Definitely breaking principle A in Ethical Guidelines for Statistical Practice I think. NadVolum (talk) 21:13, 13 July 2024 (UTC)
- GHM hasn't mentioned that, but a Professor and Department Head of Security Studies at Marine Corps University's Command and Staff College has.VR (Please ping on reply) 18:25, 13 July 2024 (UTC)
- Re the "Ken M" theory, the GHM hasn't mentioned anything about a delayed process like that. It seems reasonable for Wyner to take the published data at face value, and not speculate about hypothetical GHM processes that might explain errors in the published dates. Perhaps he could have asked GHM for an explanation, but he might have no means of contacting them. — xDanielx T/C\R 21:44, 10 July 2024 (UTC)
- Largely unanswered only in the sense that the answers are not in a form that can be used on Wikipedia. A responder to Pachter's blog for instance gave a perfectly good possibility in that the checking of deaths was done in separate batches for men, women and children and that would completely explain the strong negative correlation. This should have been an obvious possibility for Wyner but he didn't check and said 'This lack of correlation is the second circumstantial piece of evidence suggesting the numbers are not real'. What statistician finds possible problems and just writes it up that way without checking? And do you really think the GMH would put out figures like that if they were really making them up and according to Wyner had the overall figures going up too smoothly? That's just daft. NadVolum (talk) 08:58, 10 July 2024 (UTC)
- Should not include a mention of the Wyner article. Seems WP:undue to include his analysis if its a minority view published in Tablet (magazine), a conservative leaning publication that isn't cited by major news articles nor is it peer-reviewed. At the very least, the synthesis of such datasets like this is clearly ideologically minded. Wyner's previous publications in the magazine includes an article against child booster vaccination of COVID-19, despite his lack of public health background, and a study decrying California ethnic studies requirements by using statistics. This is clearly a guy who uses his background in stats to push a specific agenda, and while i'm not necessarily against including opinions of GHM as inaccurate, the single opinion of one guy, uncorroborated by any other statistical analysis by other stats people, should not be used extensively. Bluethricecreamman (talk) 01:09, 10 July 2024 (UTC)
- Unfortunately there aren't any peer-reviewed studies about GHM data. I don't think Wyner's article about COVID boosters for kids was particularly extreme? Most states ultimately didn't end up requiring boosters in schools. Opposing California's new ethnic studies requirement also doesn't seem unreasonable - after all most states have no such requirement. Not sure what the overarching agenda would be - he seems to just enjoy commenting on politics and maybe arguing somewhat contrarian (but not fringe) views. — xDanielx T/C\R 05:57, 10 July 2024 (UTC)
- His climate denial does go well into fringe territory though. Here's what Michael E. Mann had to say about the Tablet article [12]. I did not include it in the list of academics rejecting it as he leaves that to the reader rather than saying that himself and it is more an attack on Abraham Wyner himself. NadVolum (talk) 08:12, 10 July 2024 (UTC)
- Mann is a biased party here, since Wyner's testimony was an obstacle in his litigation, which won ~$1m damages but is being appealed. I'm not sure he's a climate denier - his (peer reviewed) paper was arguing about degrees of uncertainty in models, not denying that there are serious risks. Supposedly his nickname also appears on a climate denial blog, but unclear what his involvement was. — xDanielx T/C\R 22:07, 10 July 2024 (UTC)
- It is WP:UNDUE in an article about the GHM to include the contrarian opinion of Wyner if it is also undue to include his statistics about why vaccine hesitancy is valid for COVID-19 boosters. Public health statistics are significantly more complicated than Wyner's article lets on. And Wyner's views on ethnic studies being dressed up with statistical arguments is still an opinion piece at the end of the day. I suggest that all his work on Tablet magazine is necessarily an opinion piece.
- He is not an expert in public health, nor is he an expert in international affairs. He is a statistic professor at Wharton who does primarily sports analytics. His usage of statistics to dress up arguments should not disguise the fact that his piece is still an opinion piece, and suggesting that his work is useful or central in the realm of international affairs/public health around the analysis of the gaza health ministry remains a stretch.
- If we do cite him, we should not spend much time citing him, and include him among other opinions about the reliability of GHM with the correct attribution of who he is i.e. "Wyner, a professor of data science specializing in sports analytics, has claimed to have used statistics to identify abberations in Gaza Health Ministry." Bluethricecreamman (talk) 14:55, 10 July 2024 (UTC)
- I agree these are opinion pieces, but WP:RS doesn't prohibit or even discourage mentioning opinion pieces, it just requires attribution which we have.
- Ultimately if Wyner is the WP:BESTSOURCE for a significant minority view, WP:UNDUE essentially requires us to include it. If there's a competing candidate source for the view (possibly the article Alaexis mentioned?), then it could make sense to get into assessments of each professor's credibility, but otherwise it doesn't seem to matter if were some weaknesses in his unrelated past works. — xDanielx T/C\R 06:00, 11 July 2024 (UTC)
- I'm afraid that if it is Wp:OPINION, but he attempts to use armchair statistics to justify his argument, despite having no background in international affairs/public health/mortality, that we will state Wyner as some far more authoritative figure than he is. And then, when the inevitable calls for including the counterarguments debunking his stats occurs, we end up including an entire section on his stats.
- Then the article becomes Gaza Health Ministry + some sports data analytics guy's opinion about why data is bad and other people disagreeing with him
- I argue his opinion is not a significant minority view. It is the along the same line of criticism of "GHM is unreliable because its propaganda" opinion, and the sources that cite him do the same. We can include that opinion, I think, without having to go into the deep analysis of why this opinion vs that opinion blah blah internet argument; we are not here to document every turn of the argument about some sports data analysis guy beefing with other stats guys who also aren't internation affairs folks.
- It's like the same as trying to argue to include his Tablet argument about child camp covid boosters in a COVID-19 article, we should recognize him as an armchair statistician pursuing a specific argument and justifying it in hindsight with his own hobby data analysis. Bluethricecreamman (talk) 18:22, 12 July 2024 (UTC)
- His climate denial does go well into fringe territory though. Here's what Michael E. Mann had to say about the Tablet article [12]. I did not include it in the list of academics rejecting it as he leaves that to the reader rather than saying that himself and it is more an attack on Abraham Wyner himself. NadVolum (talk) 08:12, 10 July 2024 (UTC)
- Unfortunately there aren't any peer-reviewed studies about GHM data. I don't think Wyner's article about COVID boosters for kids was particularly extreme? Most states ultimately didn't end up requiring boosters in schools. Opposing California's new ethnic studies requirement also doesn't seem unreasonable - after all most states have no such requirement. Not sure what the overarching agenda would be - he seems to just enjoy commenting on politics and maybe arguing somewhat contrarian (but not fringe) views. — xDanielx T/C\R 05:57, 10 July 2024 (UTC)
- I'd prefer not to have the Wyner article but it has been referenced in some other publications and Wyner has appeared in interviews on Youtube and there's various social media discussions and blogs mentioning it. Basically it has been well publicized even if it hasn't really made it as reliable news. I woud be happy enough if there is some mention but we should include expert opinion on it. If experts with Wikipedia articles dismiss it they shouldn't be required to spend time writing an essay pointing out the problems instead of saying as one does that it is one of the worst abuses of statistics they've ever seen. NadVolum (talk) 09:23, 10 July 2024 (UTC)
Four academics who have entries on Wikipedia including two of statistics have dismissed the article
Unless there are some equally notable supporters of Wyner's statistical analysis, then it would seem that the analysis is RS discredited for our purposes. Selfstudier (talk) 13:58, 10 July 2024 (UTC)
- The RfC statement is non-neutral as there were also academics who did not dismiss Wyner's analysis [13]
- Those academics did not say anything about Wyner's article which indicated they did anything other than read the headline. This is the entire amount they said about it "Several of these oddities have recently been noted elsewhere, for instance by Salo Aizenberg, Gabriel Epstein, and Abraham Wyner." Also none of them are mentioned in Wikipedia or seem to have any particularly relevant skill and like Wyner's article it was published in a magazine promoting Israel. I'll have a proper read through that but I'm having difficulty because they mentioned registered and unregistered and I'm not exactly sure what they mean as they've made up their own terms. It may be that by registered they mean identified or it may mean passed through the hospital system and unregistered may mean recorded minus identified or it may mean recorded by media sources. There is a particular problem in that many identified deaths may not actually be amongst the recorded deaths - they're done by different systems. NadVolum (talk) 23:53, 10 July 2024 (UTC)
- Re "registered", it seems like a best-effort translation or interpretation of some info from (Arabic) PDFs the GHM released. Registered would include deaths recorded by a hospital or morgue; unregistered would include media sources and the Google form. See also this WaPo opinion, which gives similar numbers albeit using different terminology. — xDanielx T/C\R 04:43, 11 July 2024 (UTC)
- That shows the problem with people just making assumptions about terms. THe forms can only be used to identify a dead person, they can't be used to increase the recorded number of deaths - and the forms can identify a person whose death has not been recorded but is missing, probably under the rubble. The reliable media reports increase the number of recorded deaths. The unidentified in the GHM figures is recorded minus identified which is a bit fuzzy in meaning but will have a meaning when all the missing are found. An added complication with using ratios with the identified ones is that a major use of the forms is widows recording the death of their husband so they can get state aid. One needs to be careful with what is being referred to. NadVolum (talk) 06:56, 11 July 2024 (UTC)
- You're saying the Google form submissions aren't counted toward GHM's casualty count without another corroborating source? That would seem to contradict the WaPo opinion above, which says 11% of the count was self-reports (as of May 3). — xDanielx T/C\R 00:26, 21 July 2024 (UTC)
- They don't count towards the totals. They do however count towards the identified numbers. The coorrobration isn't someone saying they've seen the body, just someone like at the lawyers saying yes you are the person making the statement and I believe it to be credible correct sort of thing. NadVolum (talk) 18:07, 25 July 2024 (UTC)
- Just to be absolutely clear, an identification form never increases the total figure. The corroboration I talked about is simply an attestation that the form has been filled out in good faith. Forms reduce the number the GHM issues of unidentified casualties, which is total recorded deaths minus identified deaths. NadVolum (talk) 21:49, 25 July 2024 (UTC)
- They don't count towards the totals. They do however count towards the identified numbers. The coorrobration isn't someone saying they've seen the body, just someone like at the lawyers saying yes you are the person making the statement and I believe it to be credible correct sort of thing. NadVolum (talk) 18:07, 25 July 2024 (UTC)
- You're saying the Google form submissions aren't counted toward GHM's casualty count without another corroborating source? That would seem to contradict the WaPo opinion above, which says 11% of the count was self-reports (as of May 3). — xDanielx T/C\R 00:26, 21 July 2024 (UTC)
- It looks like they're just talking about recorded deaths which makes things easier. I'd prefer if an expert had a look at it but it seems to me to not have anywhere near so much of the egregious nonsense of Wyner's 'analysis'. The 70% does look like boilerplate the ministry just kept sticking in instead of checking and they should have checked that. The 92% women and children they derived from assuming the 70% was accurate is very sensitive to the figure, if it was 65% you get 75% women and children in the media figures which is what you'd get from indiscriminate bombing. The 10 men they then come to is more concerning - I think at the time that was done they were only using first aiders and ambulance crews for 'media' and it implies they took dead men to hospital but left many of the women and a large proportion of children at the scene. If children's deaths were properly recorded we'd expect them to be about twice that of women and it is known from other wars that children are more liable to die than men or women in the same circumstance when bombed. I read somewhere about them including two batches once and having to remove the figures, that's probably one lot of the resurrected men. They had four people originally doing the statistics in Gaza, one of them was killed and the rest had to move south around then. Israel had to resurrect 200 people after October 7 and it had nothing like the mess. NadVolum (talk) 09:31, 11 July 2024 (UTC)
- That shows the problem with people just making assumptions about terms. THe forms can only be used to identify a dead person, they can't be used to increase the recorded number of deaths - and the forms can identify a person whose death has not been recorded but is missing, probably under the rubble. The reliable media reports increase the number of recorded deaths. The unidentified in the GHM figures is recorded minus identified which is a bit fuzzy in meaning but will have a meaning when all the missing are found. An added complication with using ratios with the identified ones is that a major use of the forms is widows recording the death of their husband so they can get state aid. One needs to be careful with what is being referred to. NadVolum (talk) 06:56, 11 July 2024 (UTC)
- Wyner's analysis should be mentioned as it has been covered by RS [14] [15] [16]. Alaexis¿question? 22:16, 10 July 2024 (UTC)
- They seem to be opinion but as I said there's enough about it anyway I think to justify inclusion. NadVolum (talk) 23:53, 10 July 2024 (UTC)
- Oppose inclusion as it is indeed fringe and beyond that just blatantly wrong. Wyner claims that there isnt a variance in deaths that would be expected, but there is. The data is cherry picked, and beyond that much more higher quality sources have nearly uniformly shown his argument to be wrong. And even if it were not, it has become outdated. nableezy - 21:42, 11 July 2024 (UTC)
- Also would like to note this article is not about statistics in the current hostilities, it is an article on the health ministry itself. The claim that this is a significant minority view on that topic makes close to no sense. I don’t think it’s even a significant minority view on the ministry’s stats for the current war, but that’s a subtopic of this topic anyway. So even if it were a significant view on the stats from this war it is wildly undue for this article. nableezy - 02:12, 13 July 2024 (UTC)
- Oppose inclusion as grossly undue fringe and opinion covered largely only in other opinion and unreliable sources. It has been roundly disassembled multiple times. It could potentially find a home at Misinformation in the Israel–Hamas war based on the counter analysis, but not here. Iskandar323 (talk) 16:34, 12 July 2024 (UTC)
- Unfortunately like most that one would be unlikely to pass as a reliable source - it seems to be by a student in a student magazine. That's why I started a RfC as there's objections to including experts who just say Wyner's article is rubbish rather than giving detailed reasons for each point. NadVolum (talk) 22:43, 12 July 2024 (UTC)
- And yet, even so, it still provides a far more methodical and serious assessment and makes for better reading than the original piece. Also, for the purposes of disassembling an opinion, it's perfectly adequate. Iskandar323 (talk) 05:28, 13 July 2024 (UTC)
- I have seen one or two similar that take apart Wyner's analysis in excruciating detail and while not passing the RS bar, are quite convincing nonetheless. Selfstudier (talk) 07:34, 13 July 2024 (UTC)
- And yet, even so, it still provides a far more methodical and serious assessment and makes for better reading than the original piece. Also, for the purposes of disassembling an opinion, it's perfectly adequate. Iskandar323 (talk) 05:28, 13 July 2024 (UTC)
- I'm fine with moving it to Misinformation in the Israel–Hamas war or elsewhere, but lets not remove it entirely.VR (Please ping on reply) 18:28, 13 July 2024 (UTC)
- Unfortunately like most that one would be unlikely to pass as a reliable source - it seems to be by a student in a student magazine. That's why I started a RfC as there's objections to including experts who just say Wyner's article is rubbish rather than giving detailed reasons for each point. NadVolum (talk) 22:43, 12 July 2024 (UTC)
- Mention both Wyner's allegations and all the rebuttals. I'm astounded that XDanielx tries to remove even the most basic mathematical rebuttals against Wyner and then claims "
Wyner's questions remain largely unanswered
". We are in mathematical territory here, and wikipedia should not shy away from detailed statistical analysis sourced to recognized experts. Random math article with proofs.VR (Please ping on reply) 18:15, 13 July 2024 (UTC)- I never suggested removing Pachter's blog, and I don't see what else there is that could be considered a "mathematical rebuttal". I don't really consider the unfounded speculation by blog commenter Ken M to be an "answer". — xDanielx T/C\R 16:35, 19 July 2024 (UTC)
- Oppose inclusion There are multiple reliable sources saying GHM is reliable and Wyner in effect saying it isn't, in addition his analysis has been questioned by reliable sources. Selfstudier (talk) 18:21, 13 July 2024 (UTC)
- The fact that RS have mentioned it is reason to include it. We frequently have articles on information debunked by RS (COVID-19 misinformation etc).VR (Please ping on reply) 18:27, 13 July 2024 (UTC)
- Just because something is verifiable does not mean that it should be included, right? Selfstudier (talk) 18:29, 13 July 2024 (UTC)
- Inclusion is subjected to consensus, yes. But I strongly think inclusion here is a good thing. We can have it at a separate article, but there has been much discussion on GHM reliability. Rather than just saying "GHM is reliable", we should include the supporting mathematical discussions, in sufficient mathematical detail. We do that at all our other math articles.VR (Please ping on reply) 18:47, 13 July 2024 (UTC)
- We can say that GHM is reliable without the need for any mathematical discussions because there are a ton of sources saying they are, against which Wyner's analysis rather pales. Selfstudier (talk) 18:54, 13 July 2024 (UTC)
- I think at this point a WP:SPINOFF is definitely in order. So we can leave behind a summary here that GHM is reliable and keeping the scientific analysis on the sub article. Thoughts @Bluethricecreamman and @NadVolum?VR (Please ping on reply) 19:03, 13 July 2024 (UTC)
- One could very readily argue that the very fixation on the reliability of the GHM here is itself a POV one – and something of a red herring in the conflict, just like so much of the other disinformation. It doesn't need a spin-off; it just needs discussing on the misinformation page. Iskandar323 (talk) 20:15, 13 July 2024 (UTC)
- Looking at the longer term such a spin off article would just be a stub that festers on Wikipedia. It won't take up that much space in this article or/and the misinformation article. NadVolum (talk) 20:43, 13 July 2024 (UTC)
- I'm ok with moving it there. I'm just strongly opposed to deleting it from wikipedia entirely.VR (Please ping on reply) 19:13, 16 July 2024 (UTC)
- I think at this point a WP:SPINOFF is definitely in order. So we can leave behind a summary here that GHM is reliable and keeping the scientific analysis on the sub article. Thoughts @Bluethricecreamman and @NadVolum?VR (Please ping on reply) 19:03, 13 July 2024 (UTC)
- We can say that GHM is reliable without the need for any mathematical discussions because there are a ton of sources saying they are, against which Wyner's analysis rather pales. Selfstudier (talk) 18:54, 13 July 2024 (UTC)
- Inclusion is subjected to consensus, yes. But I strongly think inclusion here is a good thing. We can have it at a separate article, but there has been much discussion on GHM reliability. Rather than just saying "GHM is reliable", we should include the supporting mathematical discussions, in sufficient mathematical detail. We do that at all our other math articles.VR (Please ping on reply) 18:47, 13 July 2024 (UTC)
- Just because something is verifiable does not mean that it should be included, right? Selfstudier (talk) 18:29, 13 July 2024 (UTC)
- The fact that RS have mentioned it is reason to include it. We frequently have articles on information debunked by RS (COVID-19 misinformation etc).VR (Please ping on reply) 18:27, 13 July 2024 (UTC)
RFC of interest
editPlease see Wikipedia:Neutral point of view/Noticeboard#RFC - Gaza Health Ministry qualifier. ScottishFinnishRadish (talk) 12:54, 15 July 2024 (UTC)