Wikipedia:Reliable sources/Noticeboard/Archive 428

This is an archive of past discussions on Wikipedia:Reliable sources. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current main page.

Archive 425

Archive 426

Archive 427

→

Two branches of the same tree: OW-JP-AH and WCE-WRD/WCD

Operation World, Joshua Project and Asia Harvest

Asia Harvest (https://www.asiaharvest.org/) is an American Christian missionary organisation focusing on Asia and especially on China. They produce extremely detailed (and overestimated) fantasy statistics about Christians for each one of the smallest administrative divisions of the country.

Let's take, for example, the purported 2020 statistics for Shanghai (https://www.asiaharvest.org/christians-in-china-stats/shanghai). As you can see, they extrapolate absolute numbers on the basis of the very same percentage values for the total population numbers of most of the districts, and then the resulting numbers are divided according to the various statistical subcategories. Amongst the numbers in the tens of subcategories, they cite sources for only three of them, and they are some journals (probably missionary journals) dated to 1990, 1991 and 1992, while the general data are presented as being dated to 2020. The source for some of the totals is, otherwise, Operation World (https://operationworld.org/), "the definitive volume of prayer information about the world", associated with the Joshua Project, which is already classified as unreliable in the WP:RSP list.

I propose that Asia Harvest and Operation World be added to the Joshua Project entry in the WP:RSP list. Besides, on the Wikipedia article about Operation World it is written that the subject is related to the World Christian Encyclopedia, the predecessor of what is now published as the World Christian Database and World Religion Database, themselves thoroughly discussed in 2018 and 2022-2023, and listed in WP:RSP. Æo (talk) 18:09, 1 January 2024 (UTC)

I ping Erp who raised doubts about the extreme precision of WCD/WRD data in the abovementioned 2022-2023 discussion, since the same argument applies to this case. Æo (talk) 18:16, 1 January 2024 (UTC)

Besides, on the Wikipedia article about Operation World it is written that the subject is related to the World Christian Encyclopedia, the predecessor of what is now published as the World Christian Database and World Religion Database, themselves thoroughly discussed in 2018 and 2022-2023

The meaning of this point is somewhat lost on me. According to the close of the linked 2023 discussion, There is no consensus to deprecate these sources (bolding added). If consider Asia Harvest or Operation World is/are affiliated with/comparable to the WCE as a source, that would suggest not deprecating them, but instead merely advising editors to use them with prudence while favoring, where available, stronger, more certainly reliable sources. P-Makoto (she/her) (talk) 18:43, 1 January 2024 (UTC)

They are clearly not the very same as the WRD/WCD (which is nonetheless questionable, and this is why it is in the perennial sources' list and the closing statement also says that there is rough consensus to attribute it and prefer better sources), at least according to what I have been able to find, although they cross-reference to each other (it is unclear to what extent). Asia Harvest and Operation World are on the other hand directly related to the Joshua Project, which is classifed as unreliable in the perennial sources' list: The Joshua Project is an ethnological database created to support Christian missions. It is considered to be generally unreliable due to the lack of any academic recognition or an adequate editorial process. The Joshua Project provides a list of sources from which they gather their data, many of which are related evangelical groups and they too should not be used for ethnological data as they are questionable sources.. Æo (talk) 19:41, 1 January 2024 (UTC)

Addendum:

In The Ethos of Operation World we can read the following statement: We pray that these statistics and prayer points present a reasonably balanced account of what God is doing in our world and of the challenges facing us as we press on to complete the Great Commission. Apart from Operation World, only the World Christian Database/World Religions Database shares our ambition (folly?) in attempting so massive a task as compiling a comprehensive body of data relating to the world’s religions, denominations, and churches, as well as to the progress of the Great Commission.. Here, Operation World and the WRD/WCD are clearly defined as confessional, evangelical entities working together for the "progress of the Great Commission", which is unclear whether it refers to the doctrinal concept or to the American fellowship of evangelical groups which disbanded in 2020.
In this paper by the Gordon-Conwell Theological Seminary, some of whose members are also the editors of the WRD/WCD, on pp. 16-17 the methodologies of the latter are compared to those of Operation World.

Æo (talk) 20:16, 1 January 2024 (UTC)

Apologies for this, but you seem to be saying two different things simultaneously. First you say that They are clearly not the very same (bolding added); then you say they are clearly defined as confessional, evangelical entities working together (bolding added). Are they together, or are they not; and in either case, why is that a reason for depreciation of Asia Harvest (which is the source I thought was under discussion).

In any case, it is not so clear to this reader as it is to you. The Ethos statement does not seem like evidence of organizational collaboration. Rather, it reads as an observation that they share a field of study: both are attempts at compiling a comprehensive body of data relating to the world’s religions, denominations, and churches. To use another example, both Michael Burlingame and Ronald White shared the ambition (folly?) in attempting so massive a task as narrating the life of Abraham Lincoln in single-volume biographies. But they were not collaborators.

As for the "Christianity in its Global Context, 1970–2020" document, the comparison drawn is moreover a contrast, pointing out how Operation World's definitions of "evangelical" inflate their numbers compared to the World Christian Database.

Finally, simply as a note, you emphasize connections between GCTS faculty and the World Christian Database but have left out how World Christian Database is published by Brill, an academic publisher that employs editorial and peer review. (Likewise, World Christian Encyclopedia was published by Oxford University Press, also an academic publisher that employs editorial and peer review.) That, plus their relative contemporaneity (as both were published in the twenty-first century) instills a great deal of confidence in WCD and WCE as sources.

In any case, this has been a digression. The posted discussion at hand pertains to Asia Harvest and Operation World, which have different publishers and different traits. P-Makoto (she/her) (talk) 21:45, 2 January 2024 (UTC)

It is ascertained that WCE and OW originated as two branches of the same tree, and that they maintain some connections, as hinted to in the statement above about the "Great Commission" and underlined especially in the sentence in that paper Believers in Christ from a Muslim Background which I have quoted below (17:04, January 2 addendum): There has been a long history of close collaboration and mutual sharing of information among Operation World ... and the World Christian Encyclopedia. This is what I meant, and I am still investigating to find further, clearer evidences. Besides, AH and OW appear to be related as well, given that the few references showed by AH are mostly to OW statistics, and in turn OW is clearly connected to the Joshua Project (they are authored/edited by the same person, Patrick Johnstone), which is acknowledged to be a completely unreliable source.

Amongst the many discussions about the JP, read this 2008 one, which was particularly animated (and which highlights that already back then there were strange waves of spamming of this type of sources, as I myself noticed more recently); some quotes: [JP is] a very aggressive evangelistic project. ... Linking or even mentioning this project on this kind of scale should be considered as fundamentalist Christian spam. (Jeroenvrp); All links to the Joshua Project should be deleted immediately and without question. The information on the site is often original research and totally incorrect. It is not a reliable source at all. The fact that someone can't find alternative information on Google is no excuse: get out of your chair and head to a library. (Caniago); Here is another example which illustrates the sort of disinformation they are spreading. They invented a whole range sub-ethnic groups of the Javanese ethnic group, yet there are no published academic sources (in books or peer reviewed papers) which mention these sub-ethnic groups at all. There are a plethora of other examples of their disinformation if you compare their website against reliable sources. (Caniago); The project site is not an academic source. ... The Joshua Project has an religious agenda. Anyone should agree on that. This is very clear on the site and not even that, it is also very offensive. Not only for people of these ethnic groups, but for anyone who condemn these kind of aggressive evangelisation practices. I even find it very scary how they present the data (e.g. see the column "Progress Scale"). It's like: "evangelism meets the Borg". ... The data on the Joshua Project is unreliable, like others before me have proved. ... Information from the English Wikipedia is easily translated to other Wikipedia projects. Although people who translate should double check these kind of sources, unfortunately sources like the Joshua Project are spreading like a virus to those other projects. That's why I am here now, because I noticed the Joshua Project was listed as a source on the Dutch Wikipedia and learned that they came from here. So know your responsibility! ... To conclude this: I am not accusing individual Wikipedians for "fundamentalist Christian spamming". No, what I mean that on a larger scale it's "fundamentalist Christian spamming". (Jeroenvrp); There are no cases where there Josuha project is the best source of data. A bunch of evangelical missionaries are the last people who can be trusted to present non-biased reliable ethnic data; the examples we have given proven the case. (Caniago).

Regarding the fact that some of the sources we are discussing here (WCE and its successors) have been published by renowned publishing houses, this does not make them reliable. This was already pointed out in the 2022-2023 discussion. The "peer review" and "editorial process" is very often carried out by people belonging to the very same agenda and organisations (those American evangelical organisations). Take for instance the paper Believers in Christ from a Muslim Background that I quoted below: it was written by D. A. Miller, peer reviewed/edited/co-authored by Patrick Johnstone of the WEC International, and published on the Interdisciplinary Journal of Research on Religion founded by Rodney Stark (known for his publications which were very supportive of Christianity); the journal's editorial board includes Massimo Introvigne, whose CESNUR and related publications are themselves currently listed as unreliable in the WP:RSP list (and I personally consider CESNUR, or at least some of its publications, as much more reliable than the sources we are discussing here). Regarding the fact, and the problem, that the WCE and its successors have been published as seemingly academic resources, there are some further considerations expressed in a recent critical essay which I will cite and quote in a separate section below (cf. #World Christian Encyclopedia and World Religion Database/World Christian Database). Æo (talk) 00:06, 3 January 2024 (UTC)

I would agree that CESNUR and its related publications are more reliable than the assessment listed on Wikipedia's current Perennial Sources page would suggest. I think the generally unreliable characterization is inaccurate and that the academic field of religious studies has a much more favorable impression of CESNUR than Wikipedia's Perennial Sources page does.

Patrick Johnstone was not a peer reviewer of "Believers in Christ from a Muslim Background: A Global Census". Blind peer review means the reviewer is anonymous. Johnstone and Duane Miller are listed co-authors of the paper. The two peer reviewers would have been two other scholars whose identities neither of us know.

Your impression that renowned publishing houses like Brill and Oxford University Press are somehow being subverted by a conspiracy of Evangelical authors who defy the consensus of the field of religious demography stretches this editor's credulity. The peer review process is more robust than that.

That information is shared between World and WCE does not necessarily make one unreliable merely because the other is. Different sources can use the same raw data to arrive at different conclusions, such as how WCE and Operation World arrive at quite different total numbers, projections, etc.

In any case, I think that an earlier comment in this discussion from Erp rings true: for this particular discussion, we should concentrate on whether Asia Harvest and/or Operation World are reliable. There is not a consensus between us about WCE or WCD or WRD. Maybe there can yet be a consensus between us about Asia Harvest and Operation World. P-Makoto (she/her) (talk) 00:49, 3 January 2024 (UTC)

Yes. I agree with P-Makoto: "Your impression that renowned publishing houses like Brill and Oxford University Press are somehow being subverted by a conspiracy of Evangelical authors who defy the consensus of the field of religious demography stretches this editor's credulity. The peer review process is more robust than that." There is no evidence that the process is somehow compromised and is just speculation. Borders on conspiracy theory actually. In fact they show divergence of data too per already quoted differences in numbers in the sources. They are not equivalent or the same. I also agree that we should concentrate on whether Asia Harvest and/or Operation World are reliable. Ramos1990 (talk) 01:12, 3 January 2024 (UTC)

P-Makoto, I never wrote about "a conspiracy of Evangelical authors who defy the consensus" and are trying to subvert academic publishers. Apart from this, you wrote that the Ethos statement does not seem like evidence of organizational collaboration, but the statement in the Miller & Johnstone paper clearly tells us about a long history of close collaboration and mutual sharing of information. Also re-read Erp's comment below, with an excerpt from the Operation World book (2010 edition, p. 25) telling us that ... the Joshua Project List, the World Christian Encyclopedia and a handful of other resources are at the heart of this information, which is both fuel for prayer and data for mission strategy, and on that page the discourse of the author is general, about the shared project in which OW, the JP and the WCE are all actors. In my opinion, there is enough evidence to affirm that the WCE and the OW, and their affiliated projects, are still closely related. The discussion about the WCE and its successors, however, continues below (cf. #World Christian Encyclopedia and World Religion Database/World Christian Database). Æo (talk) 18:39, 3 January 2024 (UTC)

Regarding P-Makoto, I never wrote about:

You wrote that Regarding the fact that some of the sources we are discussing here (WCE and its successors) have been published by renowned publishing houses, this does not make them reliable. This was already pointed out in the 2022-2023 discussion. The "peer review" and "editorial process" is very often carried out by people belonging to the very same agenda and organisations (those American evangelical organisations) (bolding added). P-Makoto (she/her) (talk) 19:11, 3 January 2024 (UTC)

The next question is the reliability of "Operation World", multiple editions by Patrick Johnstone and Jason Mandryk with the latest being the 7th edition, published 2010, plus a web site. It is explicitly a prayer guide and does not seem to be peer reviewed. I note in reference to the Joshua Project that Operation World's website states: "The Joshua Project is our default site for people group information." https://operationworld.org/prayer-resources/helpful-resources/ Looking at the google preview of the book has "...Joshua Project List, the World Christian Encyclopedia and a handful of other resources are at the heart of this information" Given the dependence of "Operation World" on Joshua Project a "Generally unreliable source" and lack of peer review for the work itself, I would say Operation World must also be listed as "Generally unreliable source". Erp (talk) 20:13, 1 January 2024 (UTC)

I agree, and the WRD/WCD should be re-assessed as well, given its connection with Asia Harvest/Operation World. In Darrell L. Bock (2013), The Cape Town Commitment: A Confession of Faith, A Call to Action: Bibliographic Resources, p. 32, we read: These two books come from the same stable. While up to the mid-1990s the databases behind Operation World and the World Christian Encyclopedia were virtually identical, they began to diverge in the 1990s, partly because Operation World took a more generous definition of the word 'evangelical'. In 2010, World Christian Encyclopedia said there were 300 million evangelicals worldwide, whereas Operation World said there were 550 million.... On the same page, the World Religion Database/World Christian Database and the Atlas of Global Christianity are identified as the continuations of the World Christian Encyclopedia, while The Future of Global Christianity is identified as built on the database of Operation World. Other minor publications associated with them (listed on the same page) are: World Christian Trends – AD 30-2200, World Churches Handbook, Global Religious Trends 2010 to 2020, Megatrends and the Persecuted Church, Global Restrictions on Religion, Global Pentecostalism, The New Faces of Christianity, The Next Christendom, Barna Updates (https://www.barna.org), and Global Mapping International (https://www.gmi.org). Ultimately, they are all affiliated with the Lausanne Committee for World Evangelization, the same who launched the 10/40 window concept. Æo (talk) 21:17, 1 January 2024 (UTC)

The assessment of the World Religion Database and World Christianity Database strikes me as a separate question. If they are re-assessed, I would encourage re-assessing them "upward" rather than "downward". The source you cite, Cape Town Commitment, even identifies how the two sources are different: Operate World took a more generous definition of the word 'evangelical'. In 2010, World Christian Encyclopedia said there were 300 million evangelicals worldwide, whereas Operation World said there were 550 million. You speak of WCE/WRD/WCDs' connection with Asia Harvest/Operation World; however, what seems to be demonstrated is their disconnection; if Operation World and Asia Harvest are overstating, WRD/WCD/WCE apparently are holding back in comparison. P-Makoto (she/her) (talk) 01:23, 2 January 2024 (UTC)

I don't think that the WRD/WCD should be re-assessed "upwards"; their problems, which are still different from those of the AH/OW discussed here, were pointed out and thoroughly discussed with extensive quotes from critical sources in the specific 2022-2023 discussion. AH/OW and WCD/WRD are ultimately two branches of the same tree, dedicated to "the progress of the Great Commission" (cf. above), and this does not mean that if one of the two branches is unreliable the other is reliable, and vice versa. Both of them have problems, albeit differentiated. Æo (talk) 01:48, 2 January 2024 (UTC)

There appears to not be consensus between us. The specified 2022-2023 discussion also had extensive references from laudatory sources which reviewed the Encyclopedia positively. I developed an impression that the listing of WRD/WCD/WCE as "additional considerations" may have been excessive and not the right call.

But that would be a discussion different from that of the present one about Asia Harvest. P-Makoto (she/her) (talk) 03:19, 2 January 2024 (UTC)

I think we should concentrate on whether Asia Harvest and/or Operation World are reliable. Given that Operation World depends on the already listed as generally unreliable, Joshua Project, and Asia Harvest depends on Operation World that both should also be listed as generally unreliable. In addition neither seem to be peer reviewed. Does anyone disagree? Erp (talk) 03:38, 2 January 2024 (UTC)

Before leaping to "Generally Unreliable", may I ask whether "Additional Considerations" would be appropriate, and if you do not think so, why not?

I would note that peer review, while a gold standard, is not Wikipedia's only standard. Many sources subject only to editorial review and not peer review (newspapers, magazines, nonfiction books published with non-academic but still reputable presses) are accepted on Wikipedia, so the lack of peer review is not itself necessarily a point against. P-Makoto (she/her) (talk) 06:19, 2 January 2024 (UTC)

P-Makoto, True. Your observations are accurate with respect to the additional comments you have brought up. Indeed the jump to generally unreliable is why the RFC for WRD/WCD/WCE failed depreciation petty badly across the board. The academic sources did not support such a claim. Context matters to what Asia Harvest is being used on. Also numbers on China are hard to pin down. All polls are estimates for that. Ramos1990 (talk) 08:20, 2 January 2024 (UTC)

I agree with Erp that AH/OW should be classified as "generally unreliable" given the precedent represented by the related Joshua Project. The latter was the subject of eleven discussions on this noticeboard, and it was decidedly assessed as unreliable; just read how editors commented here and here, for instance: ...some argue based on the idea that they wouldn't have any reason to give inaccurate figures. This isn't a useful argument. There's also strong opposition to using them as a source. According to their list of data sources, a solid majority of their sources are just other evangelical groups... They shouldn't be ranked beside census counts as equivalent... They should be considered unusable due to a lack of verifiable methodology and recognition for statistical or academic contribution, even when setting aside all questions of advocacy and bias. (Elaqueate); We have no idea where they get their data, it's not part of their primary mission, and there's no significant penalty to them for errors, so I see no reason to consider them as a reliable source for population statistics. (Mangoe); I looked at the source, and I believe you. It's a hobby site by three random religious enthusiasts. Certainly not a reliable source for population data. (Alsee). Regarding the use of non-academic sources ("newspapers, magazines, nonfiction books published with non-academic but still reputable presses"), P-Makoto, yes, I think they should be eschewed and I always try to eschew them when I contribute to Wikipedia. Besides, other considerations apply in this specific case, given that we are dealing with a field of information, statistics, for which there are official censuses and statistical institutions which provide "hard data" — i.e. precise numerical results which constitute "facts" subject to minimal interpretation —, and even in the case we need "soft data" — i.e. unofficial and not always accurate data —, there are still impartial and reliable survey agencies to rely upon. In said field of information, we do not need WP:SPECULATIONs produced by organisations with blatant agendas of evangelism, proselytism or propaganda through unclear methodologies (in our case the methodologies are declared, indeed: word of mouth from priests, pastors and other church staff). Æo (talk) 14:08, 2 January 2024 (UTC)

Addendum: Here other users expressed other clear evaluations of the quality of the JP: Religious advocacy group, cites unreliable data sources. (PaleoNeonate); Publishes false or fabricated information, and should be deprecated. You cannot trust any of that website's claimed population numbers for ethnic groups even to an order of magnitude. (anonymous IP); Very obviously unreliable. Attempting to use it as a source is absurd. (Tayi Arajakate). The use of the Joshua Project on Wikipedia even caused the creation of an article about a non-existing ethnic group: Wikipedia:Articles for deletion/Jawa Pesisir Lor. Æo (talk) 15:06, 2 January 2024 (UTC)

Regarding the use of non-academic sources ("newspapers, magazines, nonfiction books published with non-academic but still reputable presses"), P-Makoto, yes, I think they should be eschewed and I always try to eschew them when I contribute to Wikipedia.

As individual editors, we all I suppose have the option to hold ourselves to higher standards than Wikipedia's; however, it is not consensus to, as a project, eschew newspapers, magazines, and nonfiction books published with non-academic but still reputable presses for being such. P-Makoto (she/her) (talk) 21:29, 2 January 2024 (UTC)

Looking at https://www.asiaharvest.org/christians-in-china-stats/china Multiple sources, I initially thought the TSPM and CPA figures were accurate since they are possibly official government sources (these are registered and recognized churches) except notes 3 and 4 indicate that the registered protestant number is from a 2010 survey that found 23 million registered protestants and that the numbers were adjusted to include non-adults and presumably the decade since. The number has been adjusted to 39,776,275 for 2020. In addition the table apparently took the 2010 Operation World figures of 86,910,600 protestants in 2010 (unregistered House Church and TSPM) and apparently projected forward to 2020 and got 109,650,630 (split between the 39,776,275 registered and 69,874,355 unregistered (note the increasing specificity during the data manipulation). I decided to look at what might be the overall source "2020, Hattaway, The China Chronicles, no page number given" which seems to be a 7 book series "The China Chronicles" by Paul Hattaway and published, by as far as I can see, "Asia Harvest" an organization Hattaway co-founded with his wife. I'm guessing he or his organization is also responsible for this table published on their website. Both count as self-published and not at all peer reviewed. They might accurately cite other sources. Erp (talk) 19:37, 1 January 2024 (UTC)

Here we can read that the book Operation China by Paul Hattaway has <...a foreword by Patrick Johnstone, author of the best-selling Operation World, who "I have relied much on the information in 'Operation China' during compilation of the section on China for the latest edition of 'Operation World'. May this unique book go a long way to focus prayer on the need for the gospel among these peoples.'>. Patrick Johnstone is mentioned in your comment above (20:13, January 1). AH and OW are definitely related. Æo (talk) 02:00, 2 January 2024 (UTC)

I found 54+ references to "Asia Harvest" in Wikipedia. A lot have to do with descriptions of people/languages where Asia Harvest in turn is citing another source (I suspect the "Encyclopedic Dictionary of Chinese Linguistics" for at least some which is a 1991 work in Chinese [Zhongguo yu yan xue da ci dian 中国语言学大辞典]). My guess is that Asia Harvest was used by wiki editors because it has translated some of the information into English. I suspect editors would be better off for a comprehensive work relying on Ethnologue (which has some faults but is generally accepted by scholars) though it does require a subscription. Glottolog is also useful especially for references to works on a language (less so for numbers of speakers).

Operation World is also cited (oddly enough mostly in articles about Baháʼí such as Baháʼí_Faith_in_Nigeria) which has "Estimates of membership vary widely - a 2001 estimate by Operation World showed 1000 Baháʼís in 2001 while the Association of Religion Data Archives (relying on World Christian Encyclopedia) estimated some 38,172 Baháʼís." Another source had about 15,000 in 2000 (Lee, Anthony A. (2011). The Baha'i faith in Africa: establishing a new religious movement, 1952-1962. Studies of religion in Africa. Leiden ; Boston: Brill. ISBN 978-90-04-22600-5., page 107, itself citing an unpublished article). I'm inclined to go with the peer reviewed book. Erp (talk) 05:05, 2 January 2024 (UTC)

Duane Miller's Believers in Christ from a Muslim Background: A Global Census (n.b. edited by Patrick Johnstone, WEC International), questioned on this noticeboard in 2017, was built on Operation World and Joshua Project data. On pp. 3-4 we find further details about their parent organisations (as of 2015) and author: The results of this massive, multidecade data collection effort were eventually made available in the form of the religious data on the Operation World website, which is hosted by Global Mapping International, and the ethnolinguistic data on the interactive website of the Joshua Project, for which Johnstone was a senior editor. Therefore additional details on the sources of our information can be found at the website of the Joshua Project, which is currently managed by the U.S. Center for World Missions.. If my understanding is correct, based on our previous findings, Johnstone was ultimately behind both Operation World and the Joshua Project. Æo (talk) 16:23, 2 January 2024 (UTC)

Addendum: Let's keep this secondary, as suggested above, but on the same p. 3 we read: There has been a long history of close collaboration and mutual sharing of information among Operation World, the Summer Institute of Linguistics, and the World Christian Encyclopedia.. Æo (talk) 17:04, 2 January 2024 (UTC)

The Miller and Johnstone source clearly supports that Operation World is a reliable source by the way. Ramos1990 (talk) 20:08, 2 January 2024 (UTC)

Johnstone is the editor/author of the OW and JP themselves. Therefore, that paper is a completely unreliable source. Besides, in the 2017 discussion one of the commentators correctly pointed out that the study misused the word "census" (which has a very precise meaning) in its title, misleading readers to think that the statistics presented were really from a census, when they were not: The author declares that he has published "a global census": the problem is that a census is "an official enumeration of the population, with details as to age, sex, occupation, etc.". So no, it's clearly not a census of any kind. Far from that. (AlessandroDe). Æo (talk) 19:46, 3 January 2024 (UTC)

Heartily disagree, you can't point to a walled garden as evidence of reliability. Horse Eye's Back (talk) 05:27, 4 January 2024 (UTC)

Agreed with AEo and Erp, these publications should really just be grouped together in one GUNREL entry here. They're all interdependent and interrelated using the same evangelical propagandizing. JoelleJay (talk) 22:08, 4 January 2024 (UTC)

@JoelleJay: I agree that they should be grouped as a single WP:GUNREL entry. Would you also support a deprecation? I decided to open this discussion since a few days ago I noticed that the OW-JP, through AH, is still being spammed throughout various articles without attention to its problematic nature and classification as unreliable in the perennial sources' list. This has been ongoing since the 2000s, unfortunately, and even on other Wikipedias, as the user Jeroenvrp from the Dutch Wikipedia complained in the comment quoted above from 2008: unfortunately sources like the Joshua Project are spreading like a virus. This is why I think that, perhaps, it is time for the further step of deprecation.

Also, what is your opinion about the related World Christian Encyclopedia and World Religion Database/World Christian Database discussed below? Æo (talk) 01:43, 5 January 2024 (UTC)

Grouping all three together under a single GUNREL entry seems straightforward enough, its a compact ecosystem and all of them are generally unreliable. Horse Eye's Back (talk) 04:42, 6 January 2024 (UTC)

I think it is worthwhile to open a new RfC for the deprecation of the OW-JP-AH datasets. Back in 2021, Alaexis, who opened that year's RfC, which resulted in the classification of the JP as "unreliable", complained that it was still being used. Things have not changed since; just take a look at these September 2023 additions (of OW-JP data via AH), complete with a map (and this is not the only case).--Æo (talk) 21:06, 7 January 2024 (UTC)

RfC: OW-JP-AH

The following discussion is an archived record of a request for comment. Please do not modify it. No further edits should be made to this discussion. A summary of the conclusions reached follows.

Procedural close; the RfC statement was neither neutral nor brief. voorts (talk/contributions) 03:01, 17 February 2024 (UTC)

The #Operation World, Joshua Project and Asia Harvest are databases of religion demographics related to the Christian missionary movement. OW and the JP are both edited by the Christian missionary Patrick Johnstone, while AH, which reproduces OW-JP statistics for Asia and China, is edited by the Christian missionary Paul Hattaway. The JP has been the subject of more than ten discussions on this noticeboard, with almost all comments finding it completely unreliable. The latest discussion with RfC in 2021 decided its inclusion in the perennial sources' list as a generally unreliable source. Despite this, it is still widely used throughout Wikipedia (cf. 1), and its associated projects OW and AH are also widely used (cf. 3, 4), and this was already a matter of complaint in the previous discussions.

Should the JP, and its associated projects OW and AH, be WP:DEPRECATED? Answer yes or no.

Æo (talk) 17:30, 10 January 2024 (UTC)

Yes, it's time to deprecate them. The JP was categorised as unreliable in the perennial sources' list with a 2021 RfC, after more than ten discussions on this noticeboard in which comments were almost universally unanimous on the serious unreliability of the source. Despite its classification as unreliable, it, and its related projects OW and AH, continue to be used uncritically in various Wikipedia articles.--Æo (talk) 17:37, 10 January 2024 (UTC)

Yes, unreliable sources, frequently discussed. 2A02:1810:BC3A:D800:A050:6C5A:A34E:91A2 (talk) 11:30, 16 January 2024 (UTC)

Bad RfC -- also, it seems like from the timestamps that it's been snuck in through the back of an existing RfC thread. SamuelRiv (talk) 20:02, 25 January 2024 (UTC)
Bad RfC per SamuelRiv. GretLomborg (talk) 07:10, 30 January 2024 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

World Christian Encyclopedia and World Religion Database/World Christian Database

Our latest discussion about the World Christian Encyclopedia and its successors, the World Religion Database and World Christian Database, currently also presenting their statistics through the platform of the Association of Religion Data Archives, was in late 2022-early 2023. As demonstrated in the section above (see comments 20:13, 1 January by Erp; 20:16, 1 January addendum by Æo; 17:04, 2 January addendum by Æo; 18:39, 3 January by Æo), the WCE and its successors have some connection and/or collaborate and share information with Patrick Johnstone's Operation World and Joshua Project and their network (incl. Paul Hattaway's Asia Harvest, et al.), and ultimately the WCE and OW branched out around the mid 1990s from the same statistical database, and they all seem to be affiliated with the Lausanne Committee for World Evangelization (see comment 21:17, 1 January by Æo).

A new critical essay about the WCE and its successors, which adds to those already mentioned in the foregoing 2022-2023 discussion, was published right last year: Adam Stewart's Problematizing the Statistical Study of Global Pentecostalism: An Evaluation of David B. Barrett's Research Methodology, in Michael Wilkinson & Jörg Haustein's The Pentecostal World (Routledge, 2023, pp. 457-471). It criticises the methodologies of David B. Barrett, a Welsh Anglican priest and the creator of the WCE, which were used to compile the WCE itself. Todd M. Johnson and Gina A. Zurlo, who are also mentioned in the essay and are the theorists and directors of "Global Christianity" studies at the Gordon-Conwell Theological Seminary, are otherwise the continuators of the WCE in the form of the WRD/WCD.

Within the essay, the author elaborates: <... what I call the “Pentecostal growth paradigm,” initially promulgated by David B. Barrett, and now ubiquitous within the field of Pentecostal studies, as well as four common critiques of the paradigm ... the complicated typology conceptualized by Barrett in the first edition of the World Christian Encyclopedia in order to classify and measure Pentecostals around the world ... the – very limited – information that Barrett provides regarding the data collection techniques that he used to gather the data contained in the first edition of the World Christian Encyclopedia ... the construct validity threats contained within Barrett’s typology of Pentecostalism and data collection techniques, which, I argue, provide sufficient evidence to substantiate previous claims that the Pentecostal growth paradigm lacks the methodological rigor required to provide valid research results ...> (p. 457).

Other quotes:

pp. 457-458: Stewart explains that some Christian authors have pushed for: <... a trend of steadily increasing estimates of global Pentecostal adherence ranging anywhere from 250 to 694 million ... The genealogy of this authorial ritual can be traced back to David B. Barrett’s original attempt to enumerate all of the various forms of global Christianity published in the first edition of the World Christian Encyclopedia in 1982, which, he argued, revealed the substantial numerical growth of Pentecostalism between 1968 and 1981. This is confirmed by Johnson who writes that “virtually all estimates for the number of Pentecostals in the world are related to Barrett’s initial detailed work”. Barrett persisted in this project for another two decades, which was continued by his closest academic successors, namely, Todd M. Johnson and, more recently, Gina A. Zurlo, who continue to record the ostensibly boundless growth of Pentecostalism around the world, a perspective which I refer to here as the Pentecostal growth paradigm ...>;
p. 458: He explains that such a paradigm was adopted and fueled by church leaders: <... who flaunted estimates of Pentecostal growth in an attempt to legitimate their particular religious organizations, proselytistic efforts, beliefs, and/or practices. Non-Pentecostal scholars of Pentecostalism, of course, also played no small role in reifying the Pentecostal growth paradigm. Estimates of the dramatic numerical growth of Pentecostalism served “to legitimate their work among their disciplinary peers who largely understood Pentecostalism as either a social compensatory mechanism for the poor, uneducated, and oppressed or – from the opposite perspective – an oppressive form of cultural imperialism that homogenizes vulnerable poor and uneducated global populations” ...>, and explains that <Some scholars of Pentecostalism – even when sometimes citing the continually ballooning estimates of global Pentecostalism themselves – are critical of the Pentecostal growth paradigm, and, especially, of Barrett’s contribution to this discourse. In my review of the academic literature, I detect four common critiques of the Pentecostal growth paradigm. First are concerns that Barrett’s early research methodology might not have been sufficiently sophisticated to provide valid results. Second is the charge that Barrett’s use of the three waves metaphor carries an ahistorical, Americentric, and teleological bias ... Third, is a more specific critique closely related to the more general second critique, which asserts that, although the increasing prevalence of Pentecostal adherence around the world is not seriously debated by scholars of Pentecostalism, a significant portion of increasing Pentecostal growth estimates are the result of definitional sprawl rather than an increase in the actual number of adherents ...>;
p. 459: He cites, amongst others: <Allan Anderson, who has characterized Barrett’s estimates of global Pentecostalism as, variously, “wild guesses,” “debatable,” “inaccurate or inflated,” “considerably inflated,” “wildly speculative” “controversial and undoubtedly inflated,” “inflated wild guesses,” and “statistical speculations” ...>;
p. 463: <Barrett’s description of the data collection techniques that he used in order to gather the data contained in the frst edition of the World Christian Encyclopedia was incredibly short – just two paragraphs ... Another notable characteristic of the data collection techniques employed by Barrett is a very liberal approach to estimation. He wrote, for instance, “The word ‘approximately’ is the operative word in this survey; absolute precision and accuracy are not to be expected, nor in fact are they necessary for practical working purposes. This means that although the tables and other statistics may help readers who want specific individual figures, they are mainly designed to give the general-order picture set in the total national and global context. To this end, where detailed local statistics compiled from grass-roots sources have not been available or were incomplete, the tables supply general-order estimates provided by persons familiar with the local statistical situation.” Barrett even admits to extrapolating estimates of the total national populations of those Christian organizations that largely recorded only either child (e.g., Catholics who mainly record baptized infants) or adult (e.g., Baptists who mainly record confessing adults) adherents. He explained, “the missing figure … has been estimated and added either by the churches themselves or the editors.” Barrett explained, for instance, that he estimated the total number of Catholic adherents within a country “by multiplying total affiliated Catholics (baptized plus catechumens) by the national figure for the percentage of the population over 14 years old”.>;
p. 464: Stewart comments that: <... his [Barrett's] cavalier approach to data collection and estimation raise significant red flags regarding the validity of his work.>;
p. 467: <The presence of significant monomethod bias represents a catastrophic failure of Barrett’s research design, which, as a result, does not meet the minimum standards of valid social scientific research. In addition to this more fundamental construct validity threat, the first edition of the World Christian Encyclopedia also contains evidence of five other threats to construct validity relating to data collection techniques, namely, reactivity to the experimental situation, experimenter expectancies, attention and contact with participants, cues of the experimental situation, and timing of measurement.>;
p. 468, Stewart concludes: Unfortunately, the research methodology employed by Barrett – specifically his typology of Pentecostalism and data collection techniques – was simply too flawed in order to provide valid social scientific research results that can be trusted and longitudinally or geographically compared. My analysis confirms Anderson’s claim that, “Scholars should no longer assume that there are some 600 million pentecostals in the world without further qualification”>.

I have also found further older papers containing negative critiques of the WCE and its successors:

Marta Reynal-Querol & José G. Montalvo's A Theory of Religious Conflict and its Effect on Growth (Instituto Valenciano de Investigaciones Económicas, 2000). On p. 10 we read: <For the sake of comparison we also collected data directly from the World Christian Encyclopedia (WCE) using its division of groups. This data set has the advantage of being a time series, providing information for 1970, 1975 and 1980. However, as we pointed out before, this source has several shortcomings. ... Comparing to the other source of information we realize the data is biased towards Christian religion. ... The distribution of religious groups between 1970 and 1980 does not change in many countries. There are only about seventeen countries that present change in proportions. But those changes occur in countries where there is double practice and they usually imply an increase in the percentage of Christians ... Because of these reasons we take the data coming from the WCE with a lot of caution.>.
Andrew McKinnon's "Christians, Muslims and Traditional Worshippers in Nigeria: Estimating the Relative Proportions from Eleven Nationally Representative Social Surveys", Review of Religious Research, 63(2): 303-315 (Sage, 2021). In it we read: <... those assessments that make use of multiple sources of data, such as the World Christian Database (WCD), have not tended to make their calculations publicly transparent, nor clarified how they have squared the differences between contrasting indicators.>; <Figures in the most recent edition of The World Christian Encyclopedia (Johnson and Zurlo 2020) draw on figures assembled and updated as part of the World Christian Database (WCD) ... None of the particular calculations are provided, nor is there any accounting for methodological decisions in any particular case; neither transparency nor replicability are in evidence, which makes social scientific evaluation of how they reached their conclusions impossible.>; <... they also note that the Database does seem to overestimate the Christian identification, and expressed concern about what appears to be uncritical acceptance of figures provided by religious groups of their membership. With reference to one denomination in Nigeria McKinnon (2020) has recently found evidence that supports the criticisms offered by Hsu et al (2008). WCD estimates for Anglican identification in Nigeria were found to be dramatically over-estimated due to The Church of Nigeria's un-evidenced membership claims.>

--Æo (talk) 18:48, 3 January 2024 (UTC)

I see that Wilkinson & Haustein argue there are meaningful flaws in the methodology of Barrett and WCE. This criticism in a reliable source of the demographic methodology and technique is the first indication to me that there is substantial reason to be cautious about using these sources. (I remain unconvinced that the socioreligious affiliations of certain authors and editors is as much reason for alarm as you have seemed to imply.)

With Wilkinson & Haustein's detailed criticism focusing on Pentecostal demographics, would we say that additional considerations must be taken when citing WCE for specifically Pentecostal demographics?

For now, I will pause my earlier musing that WCE/WRD/WCD could be re-assessed to "Generally reliable" and would consent to them being left listed as "Additional considerations". I would suggest the description in the table be changed to emphasize that the reason for such an assessment is that reliable sources have criticized the sources on methodological grounds and that the demographic conclusions require "further qualification". P-Makoto (she/her) (talk) 10:25, 4 January 2024 (UTC)

Context matters here.

These databases are global debases from academic publishers and they provide useful data that others simply do not have. All major undertakings like this will have some methodological issues and no survey or census is immune to it. No survey or census is definitive on religion. All provide pieces of the puzzle. Two examples one on another global demographic attempt and another on a country:

Actual estimates on the "atheism" demographics show how multiple surveys do not agree on the numbers or method per each country or globally. There are many reasons why this would be the case - countries vary in understanding of religion and diverse methods each one contains. For example you would think that determining atheist rates is easy (yes/no) but its more complicated. Zuckerman's study (Cambridge Companion to Atheism) [1] states "Determining what percentage of a given society believes in God – or doesn’t -- is fraught with methodological hurdles. First: low response rates; most people do not respond to surveys, and response rates of lower than 50% cannot be generalized to the wider society. Secondly: non-random samples. If the sample is not randomly selected – i.e., every member of the given population has an equal chance of being chosen -- it is non-generalizable. Third: adverse political/cultural climates. In totalitarian countries where atheism is governmentally promulgated and risks are present for citizens viewed as disloyal, individuals will be reluctant to admit that they do believe in God. Conversely, in societies where religion is enforced by the government and risks are present for citizens viewed as non-believers, individuals will be reluctant to admit that they don’t believe in Allah, regardless of whether anonymity is “guaranteed.” Even in democratic societies without governmental coercion, individuals often feel that it is necessary to say that are religious, simply because such a response is socially desirable or culturally appropriate."

At the end he had to sift through a grip of surveys his estimate ranges from 500 million 750 million atheists worldwide from this paper. Pretty wide range. His country by country ranges are complex in p. 15-17 using numerous databases. WCE and even Operation World are used in a few without issues by Zuckerman.

Even the census data can show wide divergence with other surveys in other countries like Britain. Voas and Bruce (2004) "Research note: The 2001 census and christian identification in Britain" [2] state "Results from the 2001 population census suggest that nearly 72% of people in England and Wales may be identified as Christian. This figure is substantially higher than the proportion found by the British Social Attitudes survey and other national studies. It is also higher than the broad estimates of the size of the ‘Christian community’ previously produced by the Christian Research Association, the leading source of religious statistics in the UK (Brierley, 2003:2.2)." And even note issues with census data collection ”Another problem seems more serious. Unlike opinion polls which ask questions directly of respondents, census forms are generally completed by one individual on behalf of the entire household. There is no rule about who should take responsibility, but typically it is the head of household or at least a senior member of it."

On the WCE, The Andrew McKinnon's source does state The editors of the World Christian Encyclopedia provide reasonable methodological reflections on the different sources upon which scholars may draw in order to estimate the different religious populations of the world, as well as some of the issues that crop up as one tries to reconcile sources that disagree (Johnson and Zurlo 2020: 897–914)."

And the Marta Reynal-Querol & José G. Montalvo source does say ”For the sake of comparison we also collected data directly from the World Christian Enciclopedia (WCE) using its division of groups. This data set has the advantage of being a time series, providing information for 1970, 1975 and 1980.”

Other sources like Hsu et al. 2008 deal with methodology directly and state [3] state "Scholars have raised questions about the WCD's estimates categories, and potential bias, but the data have not yet been systematically assessed. We test the reliability of the WCD by comparing its religious composition estimates to four other data sources (World Values Survey, Pew Global Assessment Project,CIA World Factbook, and the U.S. Department of State), finding that estimates are highly correlated....Religious composition estimates in the WCD are generally plausible and consistent with other data sets."

For WRD "Given the limitations of censuses, including incomplete and irregular global coverage, potential political bias swaying the findings and the absences of many religious groups from censuses, any religious demographic analysis must consult multiple sources.[4] They state their sources which include census and surveys as well and say they are transparent to the scholarly community p. 1. Ramos1990 (talk) 15:05, 4 January 2024 (UTC)

Ramos,

Regarding your claim that these databases are global debases from academic publishers and they provide useful data that others simply do not have: this is simply false; there are statistics produced by national censuses, national statistical institutes, and independent reliable survey organisations. Regarding your claim that no survey or census is definitive on religion: censuses are official countings of the characteristics of the whole population of a country, and in the case they have any shortcomings there are other surveys produced by national statistical organisations or independent reliable survey organisations. "Independent reliable" organisations necessarily means non-confessional, non-missionary, non-evangelistic, while "survey" organisations necessarily means that they actually conduct polls among populations. The WCE/WRD/WCD, given the evidence, is neither the first, nor the second thing.

Regarding Zuckerman's study of worldwide atheism, I do not understand what it has to do with the case being discussed here: Zuckerman does not claim that his study is a census, and in any case I would not use it in Wikipedia articles in place of census statistics. Regarding Voas & Bruce's research, I also don't understand what it has to do with our case: statistics from the British Social Attitudes Survey and the Christian Research Association have never been given precedence either in Wikipedia or elsewhere over census statistics. I think that the 2001 British census finding that 72% of the population identified themselves as Christian was correct, and in any case their number has shrunken to 59% by the 2011 census, and to 46% by the 2021 census; I trust that these are the correct proportions of self-identifying Christians within the British population in the three census periods.

Regarding your excerpt from McKinnon's paper, it continues with the sentence that I already quoted above: None of the particular calculations are provided, nor is there any accounting for methodological decisions in any particular case; neither transparency nor replicability are in evidence, which makes social scientific evaluation of how they reached their conclusions impossible..

Regarding your excerpt from Reynal-Querol & Montalvo's paper, it continues with the following conclusions, also already partially quoted above: However, as we pointed out before, this source has several shortcomings. First, and probably the most important, the data does not consider the possibility of double practice, very common in Sub Saharan Africa and Latin America countries. Comparing to the other source of information we realize the data is biased towards Christian religion. A clear example is the case of Kenya in which the distribution of religions is considered to be similar to Spain or Italy. The distribution of religious groups between 1970 and 1980 does not change in many countries. There are only about seventeen countries that present change in proportions. But those changes occur in countries where there is double practice and they usually imply an increase in the percentage of Christians and a reduction in the size of animist followers. Because of these reasons we take the data coming from the WCE with a lot of caution..

Regarding Hsu et al., their full paper can be read here, it was already widely quoted in the 2022-2023 discussion and it is mostly critical of the WCE/WRD/WCD. I hope it is not necessary to repeat the same findings already explained in the 2022-2023 discussion. However, your quote is missing the following parts: ... however, the WCD does have higher estimates of percent Christian within countries. ... we find that WCD estimates of American Christian groups are generally higher than those based on surveys and denominational statistics. ... the WCD counts tiny religious minorities, classifies some Muslim groups within the neoreligionist and ethnoreligionist categories, and has higher numbers of nonreligious. (p. 680); the conclusions about correlation with other datasets: ... the WCD tends to overestimate percent Christian relative to the other data sets. Scatterplots show that the majority of the points lie above the y - x line, indicating the WCD estimate for percent Christian within countries is generally higher than the other estimates. Although the bias is slight, it is consistent, and consequently, the WCD estimates a higher ratio of Christians in the world. (p. 684); and the final conclusions: We find some evidence for the three main criticisms directed at the WCD regarding estimation, ambiguous religious categories, and bias. The WCD consistently gives a higher estimate for percent Christian in comparison to other cross-national data sets. ... We also found evidence of overestimation when we compared WCD data on American denominational adherence to American survey data such as ARIS, due in part to inclusion of children, and perhaps also to uncritical acceptance of estimates from religious institutions. ... we find the WCD likely underestimates percent Muslim in former Communist countries and countries with popular syncretistic and traditional religions. ... Data on percent nonreligious are not highly correlated among the five data sets..

Regarding the WRD's own methodology paper, it is a self-published source (n.b. Brian J. Grim is another member of the Gordon-Conwell team) and it is quite simply false that they use census statistics; their data definitely do not correspond to the statistics provided by censuses. This is obvious and anyone can demonstrate it, given that census statistics are public and accessible to anyone. Stewart's paper (p. 463) also mentions census statistics dated 1900 to the 1970s, which are obviously obsolete, and some improbable unpublished data from “unprocessed” or “incomplete” national censuses. Æo (talk) 17:00, 4 January 2024 (UTC)

Addendum: We must also remember and underline another important, critical point, which is that WCE/WRD/WCD data are speculative projections (WP:CRYSTAL) ranging from 1900 to 2050, not even survey outcomes, actually. Æo (talk) 17:28, 4 January 2024 (UTC)

P-Makoto, I would agree with your proposal to add that "reliable sources have criticized the sources on methodological grounds and that the demographic conclusions require further qualification" to the description in the table.

I would not restrict the scope of the source to Pentecostalism alone, however, owing to the fuzzy definition of Pentecostalism itself (cf. Stewart) and to the fact that its alleged 600+ million adherents, purported by the sources being discussed here, add a lot to the overall number of Protestants and Christians worldwide, and also owing to the fact that (cf. Stewart; Reynal-Querol & Montalvo; McKinnon) this demographic "athorial ritual" (as Stewart calls it) apparently originated among Anglicans and also involves the overestimation, and often self-overestimation, of the populations of other Christian denominations, including Anglicans themselves and Catholics, and therefore of Christians as a whole.

I would also agree with your proposal that the source be kept in the "additional considerations" category; otherwise, if other users think it would be more appropriate to downgrade it to the "generally unreliable" or even "deprecated" category (given the continuous spam campaigns of which they are, and will likely continue to be, the subject), I would agree with them. Æo (talk) 18:02, 4 January 2024 (UTC)

I would not agree with downgrading the sources under discussion to "Generally unreliable" or "Deprecated". I have proposed neither, and I oppose both.

You would not restrict the scope of the source to Pentecostalism. But why then does the author make Pentecostalism and the "Pentecostal growth paradigm" the scope of the argument? I would be more comfortable being cautious about how far we extrapolate those conclusions.

To clarify, I do not mean to simply add "reliable sources have criticized the sources on methodological grounds and that the demographic conclusions require further qualification" to the description in the table. Rather, I would propose replacing the present description in the table with such a sentence. The current description of editors considering the source WP:PARTISAN etc. is based on editor assessments, rather than reliable secondary sources. There is not consensus on whether or not the sources are partisan. But perhaps there can be consensus that a reliable source has said that the projections require further qualification and have methodological flaws. P-Makoto (she/her) (talk) 21:22, 4 January 2024 (UTC)

P-Makoto,

I agree the depreciation is unreasonable. But I certainly would question the Petacostal paper when Hsu et al. 2008 [5] clearly does an actual wider assessment and concludes "To address the criticisms mentioned above, we compare the religious composition estimates in the WCD to four other cross-national data sets on religious composition (two survey-based data sets and two government-sponsored data sets): the World Values Survey (WVS), the Pew Global Attitudes Project (Pew), the Central Intelligence Agency (CIA), and the U.S. State Department (State Department). In our analysis, we find support for some of the criticisms made by reviewers, but on the whole we find that WCD estimates are generally consistent with other data sets. The WCD is highly correlated with the other data sets, estimates for percent Christian, Muslim, Buddhist, and Hindu; however, the WCD does have higher estimates of percent Christian within countries." and "In sum, we find that the WCD religious composition data are highly correlated with other sources that offer cross-national religious composition estimates. For cross-national studies, the WCD may be more useful than other sources of data because of the inclusion of the largest number of countries, different time periods, and information on all, even small, religious groups.”

Additionally, Hsu 2008 also explicitly says "We ran correlations of the five data sets with each other on the percentage of adherents to the major world religions (Christianity, Islam, Buddhism, and Hinduism) as well as the nonreligious (Table 2). The WCD is highly correlated with the other four data sets, with most correlations near 0.90, which suggests that its data for percent Christian, percent Muslim, percent Buddhist, and percent Hindu are generally reliable.

Also I think that there are methodological issues with other sources like census data as is exemplified with Britain. Many countries do not even have religion questions on the census either. But no one tries to depreciate those sources. It seems too much to require more from WCE than other sources when the evidence shows it is reliable and consistent with other databases on the whole.

I think removing partisan and leaving the wording as is for in text attribution makes more sense for middle ground on the table.

Also these databases are respected by diversity of scholars and authoritative sources such as scholars of Islam (e.g. The Oxford Handbook of Politics in Muslim Societies (2022)), scholars of nonreligion / irreligion (e.g. Annual Review of the Sociology of Religion: Volume 7 (2016)), The Cambridge Companion to Atheism (2007)), Pew Research Center's uses it in own methodology and database (see Pew's methodology, The Palgrave Handbook of Islam in Africa (2020)) and also Oxford Handbook of Mission Studies (2022), Oxford Handbook of the Economics of Religion (2011), The Oxford Handbook of Religion and Europe (2022)

- Woodberry, Robert D. (2010). "World Religion Database: Impressive - but Improvable" [6] - "Despite these criticisms, we can appreciate the editors’ achievement in applying a relatively consistent methodology across the world. Furthermore, the WRD estimates are highly correlated with other cross-national estimates of religious distribution, a conclusion supported by an article by Becky Hsu and others." and also "Still, despite my criticisms, I will eagerly use these data in my research. I do not know of any better data available on such a broad scale and am amazed at the editors’ ability to provide even tentative estimates of religious distribution by province and people group."

- Brierley, Peter. (2010). "World Religion Database: Detail Beyond Belief!" [7] - "The WRD is a truly remarkable resource for researchers, Christian workers, church leaders, religious academics, and any others wanting to see how the various religions of the world impact both the global and the local scenes. It is always easy to criticize any grand compilation of statistical material by looking at the detail in one particular corner and declaring, "That number doesn't seem right." The sheer scope of this database, however, is incredible, and the fact that it exists and can be extended even further and updated as time goes forward in the framework of a respected university deserves huge applause for those responsible for it. Praise where praise is due, even if I am about to critique it."

If it is good for demographers it certainly good for Wikipedia. Ramos1990 (talk) 22:08, 4 January 2024 (UTC)

@Ramos1990 Those are old sources. Have you evidence the situation is the same 13 or more years later? Doug Weller talk 15:29, 7 January 2024 (UTC)

Hi. Well, not sure if you looked at the Handbooks I linked, but some are from the 2020s. For example here is an extract of an authoritative source The Palgrave Handbook of Islam in Africa (2020):"The 2010 Pew Report is notable in terms of its comprehensive research design. Pew utilizes demographic sources from the World Religion Database as well as extensive survey data for nineteen African states. This mixed methods design of both quantitative and qualitative sources is important because it provides a substantive way to ground truth our understanding of religious affiliations and attitudes. Published demographic data alone on religion is usually drawn from censuses which can be fraught with design problems, but Pew utilizes field tested, empirical observations." Ramos1990 (talk) 18:58, 7 January 2024 (UTC)

Do you realise that that book is discussing the Pew Research Center, and that the WRD is just mentioned in a note about the sources upon which the Pew builds its estimates? The book is neither citing the WRD directly nor discussing it. And the Pew's own criteria about its use of WRD data have already been quoted in my <19:44, 6 January> comment below. Æo (talk) 21:14, 7 January 2024 (UTC)

It does not need to. It merely incudes it as part of its positive assessment of the quantitative (WRD) and qualitative (survey data) combination. It certainly does not support your view at all (that is its worthless and useless). Pew's methodology page does not either. Ramos1990 (talk) 21:27, 7 January 2024 (UTC)

P-Makoto,

In my opinion the outcome of the previous community consensus should not be altered, and Firefangledfeather's closing summary should be kept, with its reference to WP:PARTISAN, or WP:BIASED, and WP:CRYSTAL, and just altered to add your new sentence, possibly also adding Stewart's conclusion that the source lacks the methodological rigor required to provide valid research results, and a reference to the fact that these results systematically overestimate Christianity (as found by all the critical papers quoted above) and underestimate other religions (as found by Hsu et al.). Regarding WP:BIASED, I think that it is important to keep it because in my view it is quite clear that the source is biased; for me, the relationship that it has with the OW-JP, its origins as a Christian missionary project, the fact that it is edited by the Gordon-Conwell Theological Seminary (which, by the way, is itself directly related to Billy Graham and his Lausanne Movement), are all indicators of a clear bias, and in any case, this is clearly stated by Reynal-Querol & Montalvo where they wrote we realize the data is biased towards Christian religion.

Moreover, in his essay, p. 459, Stewart further explains that Barrett directly addresses and emphatically rejects what he calls the “folly of triumphalism” ... Despite this assurance, Barrett’s occupation as a missionary, stated belief that all of the world would be evangelized by the end of the twentieth century, and, not least of all, his development of a “theology of Christian enumeration” that explains the purpose of his work as helping “the followers of Christ to discern at what points to commit their resources in order to implement their commission” serve to make this, probably, the least debatable criticism ... The particular strength of this last critique might also possibly explain why, in his recent dismissal of the critiques commonly levied against Barrett’s work, Johnson [of the GCTS] elects not to address the accusation of triumphalism..

The previous quote adds to both the problem of non-neutrality, bias, of the source and to the question of the scope of the source. In his own words, Barrett theorised a "theology of Christian enumeration", not of Pentecostal enumeration. Furthermore, Stewart on p. 460 is clear when he writes that: To describe Barrett’s enumeration of Pentecostals – let alone of Christians as a whole – in the first edition of the World Christian Encyclopedia as confusing would be a drastic understatement. Guiding the entire work is Barrett’s conceptualization of Christianity ...; and again on p. 466: Barrett's ... collection techniques in order to enumerate Pentecostals and other Christians around the world. Therefore, Barret's project affects Christianity as a whole, and not merely Pentecostalism. Stewart clept it "Pentecostal growth paradigm" apparently because such a paradigm was ... adopted and more widely disseminated by Pentecostal clergy and scholars – mostly in the Global North ... (p. 458). This is probably a reference to the OW and its affiliated networks; I remind that the book Operation World (2010, p. 25) declares that ... the World Christian Encyclopedia and a handful of other resources are at the heart of this information ....

Of course you have not proposed to classify the source as "generally unreliable" or "deprecated", I did not mean that, but I would propose it if any other users agreed, since this would help stem the ongoing spam of this source throughout Wikipedia (which has continued despite its addition to the perennial sources' list last year). Æo (talk) 00:08, 5 January 2024 (UTC)

Considering the discussion here, which is very quite long, I concur with the original 2022-2023 consensus against depreciation of these sources. They are definitely used by academic researchers and the sources presented do verify that they are good for use in Wikipedia. Robert D. Woodberry's confirmation of Hsu findings of general reliability across 4 datasets are certainly notable here as multiple sources converge on overall reliability. Keeping in mind that there are many problems with all sources including census data (WRD methodology states that only about half of the world's censuses even ask about religion and that this is declining further) certainly means that many other sources need to be used by default. This is verifiable in the US, which has nothing on religion for so many decades. And numerous other nations have removed such questions for privacy and expense reasons.

I do see room for BOTH (World Christian Encyclopedia and World Religion Database/World Christian Database) and numerous other databases to be used on Wikipedia. After all, these are all just estimates at the end and the Pentecostal and Atheism examples here exhibit the need to use multiple sources to make some sense of adherents (upper and lower estimates). I will say that polls, surveys, etc also fail to predict verifiable things like political elections [8] so I can only imagine the difficulty in religion demographics.

I think a good median on the perennial table is to keep the wording as is minus "The methodology of these sources has been questioned as WP:PARTISAN and WP:CRYSTAL." since these pass on comparison with multiple other datasets. Twice with this noticeboard by the same proposer AEO. The wording would sound neutral, very basic, inclusive, and not too specific. "Preference" does not mean "removal" or "prohibited". It allows coexistence of sources. Thus I think this is reasonable. desmay (talk) 01:28, 5 January 2024 (UTC)

After all, these are all just estimates – No, there are precise statistics from censuses and national surveys, possibly integrated by other good-quality statistics from independent neutral survey organisations, for most countries. We do not need speculative projections from non-neutral organisations of Christian evangelism. But this has already been widely discussed. The WCE/WRD/WCD are regularly spammed on Wikipedia and this causes a lot of nuisance for editors in the field of religion statistics like me, Erp and others (see here, here, etc.).

... polls, surveys, etc also fail to predict verifiable things like political elections so I can only imagine the difficulty in religion demographics – Actually, I think that a cultural identifier such as religion is much more verifiable and measurable than fleeting opinions such as political votes.

Twice with this noticeboard by the same proposer AEO – I did not open the 2022-2023 discussion myself, and, in any case, what is the problem? I also opened a discussion about WP:STATISTA last year, which resulted in its categorisation as WP:GUNREL. I read a lot, I noticed that the WCE/WRD/WCD were still being spammed throughout Wikipedia, I found new evidence of their problematic nature (the new papers presented in this discussion), and therefore I decided to open this new discussion. Æo (talk) 02:14, 5 January 2024 (UTC)

Taking in what has been said so far, at this time, for WRD/WCD/WCE, I am inclined to support user desmay's recommendations. P-Makoto (she/her) (talk) 05:17, 5 January 2024 (UTC)

Me too. I think it is a balanced recommendation. Ramos1990 (talk) 05:40, 5 January 2024 (UTC)

I have to partially correct what I expressed in my <00:08, 5 January> previous comment, given that the clause with a reference to WP:PARTISAN and WP:CRYSTAL was not present in Firefangledfeather's original closing summary, but was added by Folly Mox when they created the entry in the perennial sources' list. Yet, the new evidence (Stewart et al.) introduced with the present discussion fully justifies Folly Mox's addition, and I continue to support P-Makoto's original proposal to add a further sentence as expressed in her <10:25, 4 January> comment, and my own proposal of further additions and of category re-assessment as expressed in my <00:08, 5 January> comment, rather than desmay's proposal to return to the original closing summary. Æo (talk) 15:52, 5 January 2024 (UTC)

I placed the entry at the RSP table because I had to do some citation repair regarding these sources, I think at List of religious populations. I've found that most places these are cited in articles seem to be infoboxes and tables, which don't lend themselves easily to additional explanations about methodology etc. My sample may not reflect the total citation population.

The words I used in the entry at RSP (the only one I've ever added) were probably intended to be a summary of the close of the lengthy enormous discussion. I have no objection to the wording being changed if I've misconstrued the conversation or the close. I'm not sure if I see Firefangledfeathers bluelinked above, so courtesy ping in case they have input. Folly Mox (talk) 18:55, 5 January 2024 (UTC)

I continue to support P-Makoto's original proposal to add a further sentence

I believe this misrepresentation is accidental on Æo's part, but it is a misrepresentation of me. It was never my original proposal to add a sentence. It was my proposal that the description in the table be changed (bolding added), which I later clarified to replacing the present description in the table (bolding addeed). Any proposal which merely adds a sentence about a reliable source identifying methodological flaws while retaining the reference to WP:PARTISAN and WP:CRYSTAL would be contrary to my original position in this discussion. Such a proposal originates from someone other than myself; I suppose it would be best described as Æo's proposal, inspired by an inadvertent misunderstanding of my proposal.

Additionally, my current position (as I expressed in this diff), is support of desmay's proposal: keep the wording as is minus "The methodology of these sources has been questioned as WP:PARTISAN and WP:CRYSTAL."

I think the reference to WP:PARTISAN and WP:CRYSTAL in the table as it exists is not a consensus assessment by editors. See statements in the above discussion from myself, desmay, and Ramos1990 for examples of editor expressing consideration of the WRD, WCD, and WCE to be academically valid.

It is also not consensus that the sources are unquestionable; Æo and Erp have made clear their impression of the sources as unreliable.

"Additional considersations" seems to be an appropriate assessment, inasmuch as there is not consensus for "Generally reliable" or "Generally unreliable". P-Makoto (she/her) (talk) 21:11, 5 January 2024 (UTC)

Agree that the close of the RFC by Firefalgledfeathers did not include partisan and crystal phrase. Since Folly Mox is ok with restoring the close wording. We can remove that phrase. Ramos1990 (talk) 21:36, 5 January 2024 (UTC)

Let's not rush things. The discussion has only been open for a few days, and few people have taken part in it as of today. Moreover, Folly Mox has written that they would have no objection to the wording being changed if they had misconstrued the conversation or the close. And I think they have not misconstrued the essence of the 2022-2023 conversation. Æo (talk) 22:29, 5 January 2024 (UTC)

Folly Mox: "The words I used in the entry at RSP (the only one I've ever added) were probably intended to be a summary of the close of the lengthy enormous discussion." and pinged closer Firefalgledfeathers. Ramos1990 (talk) 00:55, 6 January 2024 (UTC)

P-Makoto,

Apologies. I originally misunderstood your use of the word "change" as implying a change by addition and not by replacement, but your clarification in your <21:22, 4 January> comment was already very clear. What I meant with my previous message is that I would support the addition of the clause formulated in your original proposal, together with other critical considerations, to the current description formulated by Folly Mox, keeping the latter as it is. Also notice that other users took part, and expressed their opinions, in the 2022-2023 discussion.

I opened the present discussion to provide further evidence, from new critical essays, about the questionability of the sources under discussion; let's focus on the merits of the new evidence provided, rather than on quibbles about the current description in the perennial sources' list. Æo (talk) 22:13, 5 January 2024 (UTC)

I have considered the new evidence presented. Seeing the new evidence presented prompted my earlier expressed decision to withdraw my suggestion to re-assess WRD/WCD/WCE as "Generally reliable" to instead support their assessment as "Additional considerations" (see my comment containing For now, I will pause my earlier musing that WCE/WRD/WCD could be re-assessed to "Generally reliable" and would consent to them being left listed as "Additional considerations". P-Makoto (she/her) (talk) 04:52, 6 January 2024 (UTC)

I would suggest:

There is no consensus on the reliability of these data sources. Rough consensus developed that the sources should be used with in-text attribution and to prefer the use of stronger sources. Some editors questioned their methodology and consider them to be partisan.

The order and tone matches many other "no consensus" RSP listings. The partisan issue was discussed more thoroughly than the point about their projections, but I wouldn't strenuously fight against including a short mention of the latter. Firefangledfeathers (talk / contribs) 01:53, 6 January 2024 (UTC)

User:Firefangledfeathers, I am not sure this captures everything form both sides because multiple editors are also not convinced of partisan and multiple editors think the methodology is appropriate and consistent with multiple databases (sources and quotes for those provided too). Even in the original RFC you closed, the same thing happened (most said "No" to depreciation 10 vs 5). Ramos1990 (talk) 02:25, 6 January 2024 (UTC)

As you look through RSP, you don't see a lot of "... but others disagreed". I think we just briefly state the most impactful concern, so that it's considered in future discussions when evaluating the source. Firefangledfeathers (talk / contribs) 02:28, 6 January 2024 (UTC)

On second thought, I'd support including a brief statement about one strength of the sources. The one that stands out the most to me in the prior discussion is that these data sources are so commonly cited by high-quality sources. Something like:

There is no consensus on the reliability of these data sources. Rough consensus developed that the sources should be used with in-text attribution and to prefer the use of stronger sources. Some editors noted that data from these sources is commonly used by high-quality publications, while others questioned their methodology and consider them to be partisan.

Firefangledfeathers (talk / contribs) 02:38, 6 January 2024 (UTC)

Thank you for that! I support this balanced version. It captures both sides and the sources that were used in this discussion and the RFC you closed. Ramos1990 (talk) 02:45, 6 January 2024 (UTC)

Thank you for taking the time to hear out some thoughts on this, and to reword the description. This version has improvement that I appreciate. I am inclined to suggest rephrasing "commonly used by high quality publications" to instead say "numerous high-quality publications"? It's a subtle difference, but there are high quality publications in topics unrelated to religious demographics that don't use these sources, so to say a source is commonly cited in high quality publications feels not quite on the mark. Saying that editors have noted that they are used in numerous high-quality publications, that seems fair and demonstrably true. P-Makoto (she/her) (talk) 04:49, 6 January 2024 (UTC)

Works for me. Not looking to make this change too soon, so you (and any others) should feel free to suggest changes or propose alternatives. Firefangledfeathers (talk / contribs) 04:52, 6 January 2024 (UTC)

P-Makoto, works for me too. Ramos1990 (talk) 06:19, 6 January 2024 (UTC)

@Firefangledfeathers: I continue the discussion, with an alternative proposal, further below. Æo (talk) 19:46, 6 January 2024 (UTC)

I'm leaning in agreement with AEo here, that the summary by FFF should be retained and editors cautioned about using these as sources. The issues over methodology are compounded by the real concern of religious advocacy/promotion/bias raised by Stewart and others. I'm also of the opinion that the very limited use of WRD by Pew Research is rather telling: they opt to cite it (as one among several databases) only in circumstances where basically no census/survey or granular data exist, rather than incorporating WRD reports into all of their estimates. JoelleJay (talk) 22:00, 5 January 2024 (UTC)

Pew uses WRD for 57 countries at least. That is a good chunk. Considering that they use "large scale demographic surveys" for 43 countries, and "general population surveys" for 42 countries, it is quite useful to complete the picture for their global estimates. Ramos1990 (talk) 22:20, 5 January 2024 (UTC)

Look at which countries though... It is a rather limited use. I would not lean on Pew to establish reliability for this source, I'd find someone who actually endorses it. Horse Eye's Back (talk) 04:47, 6 January 2024 (UTC)

Fair enough. I have multiple academic sources in purple and stuff like recent Handbooks above on it. Ramos1990 (talk) 06:24, 6 January 2024 (UTC)

But those 57 countries comprise only 5% of the population covered. JoelleJay (talk) 18:47, 7 January 2024 (UTC)

Religious demography is about studying different countries and the beliefs of the people there. I am sure you will agree that each country is a different culture with diverse beliefs and histories and that these people matter - no matter how much on a global scale they are. Approximately 4.6 billion people live in ten countries, representing around 57% of the world's population [9]. I don't think that looking at only 10 out of 232 countries are representative of the cultures of the the remaining 222. China and India alone are 38% of the world population (~3 billion). Besides if you you calculate 5% out 8 billion, its 400,000,000 people from 57 nations with diverse cultures, histories, and beliefs. That is substantial and researchers do not just throw their hands up and ignore them. Most of nonreligion research focuses in Western nonreligious populations (Europe (12%), North America (5%)), but the overwhelming majority of the nonreligious are in Asia and in particular China alone (76%) from Pew. I don't think North America should be ignored just because it is 5% of the global nonreligious population. Ramos1990 (talk) 20:17, 7 January 2024 (UTC)

It's 5% of the population because for that 5% Pew couldn't find any other sources besides WRD and some other databases. If WRD was being treated as completely reliable by Pew they would incorporate WRD data into the other 95% of their estimates. JoelleJay (talk) 23:27, 7 January 2024 (UTC)

Pew does not go by % people. It goes by Countries (25% of their countries used WRD, whereas large surveys (18%), general surveys (18%), and census (38%)). Good coverage. Each country has different understanding of religion and instruments of measure are diverse. You can read Pew's methodology to see that they say they used multiple quantitative and multiple qualitative sources for each country. Its inevitable because all sources are limited. Pew says "variation in methods among censuses and surveys (including sampling, question wording, response categories and period of data collection) can lead to variation in results". So adjustments need to be made (e.g. one source may have some data and but another source may have the missing data, but needs a third source to refine everything). In general there are 3 broad categories for religion (belief, belonging, and behavior). Some sources may have affiliation data, but not belief, or they may have belief data, but low sampling or poor wording. To keep it short, see Zuckerman in purple text, where he shows examples of massive hurdles to get a usable count on the number of atheists in any given country. Sometimes researches use more math to standardize (weighted or non-weighted). In any case, WRD is a database and it collects sources and is just one tool, among others, that researchers of every stripe do use. You can see the WRD methodology. It is available, not hidden. Also it used on continental Europe [10] by others. Hope this helps. Ramos1990 (talk) 01:41, 8 January 2024 (UTC)

In the methodology you linked previously, Pew says Together, censuses or surveys provided estimates for 175 countries representing 95% of the world’s population. In the remaining 57 countries, representing 5% of the world’s population, the primary sources for the religious-composition estimates include population registers and institutional membership statistics reported in the World Religion Database and other sources. JoelleJay (talk) 03:44, 8 January 2024 (UTC)

Pew breaks it down further "Censuses were the primary source for Pew Forum religious composition estimates in 90 countries, which together cover 45% of all people in the world. Large-scale demographic surveys were the primary sources for an additional 43 countries, representing 12% of the global population. General population surveys were the primary source of data for an additional 42 countries, representing 37% of the global population." With 57 countries for WRD, they covered more countries than large scale demographic surveys (43 countries), and general population surveys (42 countries). Population wise, large scale demographic surveys (43 countries) was 12% of the global population, which is very comparable to WRD. Of course % of people covered is irrelevant because each country has different practices and beliefs, histories (religious beliefs from China and India do not reflect most of the world despite them being 38% of the global population.) It would be odd to dismiss 57 countries out of 232, 43 countries out of 232, 42 countries out of 232, or 90 out of 232. They also state "Pew Forum researchers acquired and analyzed religious composition information from about 2,500 data sources, including censuses, demographic surveys, general population surveys and other studies – the largest project of its kind to date." Though I can see where you are coming from, I am afraid the view that there should be 1 magical super source that applies to all 232 countries is not quite possible. They had more than 10 times 232 sources analyzed and mathematically adjusted to come up with their final product for just 232 countries. Ramos1990 (talk) 05:57, 8 January 2024 (UTC)

I am in agreement with the proposal put forth by Firefangledfeathers and P-Makoto. It is neutral and on point. desmay (talk) 14:23, 6 January 2024 (UTC)

In her comments, Ramos quoted the abstract from Brierley's World Religion Database: Detail Beyond Belief! (2010) emphasising the author's seemingly positive evaluation of the source. However, reading through the essay one finds that in the conclusions the author points out that: ... This illustrates the dilemma for the compilers of the WCE and WRD. The Church of England may claim 26 million people, roughly the number living in the UK who have been baptized in the church either as infants or adults. The WRD treats this as their official source. However, not all of these now regard themselves as belonging to the Church of England and so did not tick the "Christian" box on the census form. Result? The WRD puts the Christian percent as 81 percent, the census as 72 percent, with the difference virtually entirely in the group of people who have left (as other research has shown). Which source should the WRD trust or use? This is their statistical nightmare, and the WRD in this instance opts for denominational information and does not judge between the two (though perhaps it should). This perhaps explains why some highly erudite commentators, such as Philip Jenkins, whose books on the world Christian scene have been so powerful and helpful, criticize the numbers in the WCE (and doubtless will those found in the WRD). Jenkins sometimes uses the CIA data instead, but there is no guarantee that that is more reliable.. This was written in 2010 with the data from the 2001 British census in mind; fourteen years later, things have not changed: compare WRD UK 2020 data with the 2021 UK census data.
The strength of the database, according to Brierley, merely consists in its unprecedented ... attempt on a worldwide basis to compile numbers for the different religions in a broadly compatible manner for each country.. Moreover, Brierley also concludes that: ... Christian and religious commentators have no option but to use it, despite hang-ups on definitions and individual numbers. ... These figures are not just for academic reflection and analysis but for strategic use and application. "Strategic use and application" refers to Christian mission, since Brierly is a Christian minister and/or missionary himself.
Ramos also quoted from Woodberry's World Religion Database: Impressive—but Improvable (2010); on the first page of the paper (unfortunately, I can't access the full text) we read: ... the editors seem to have constructed their estimates of religious distribution primarily from surveys of denominations and missionaries, not from censuses or representative surveys of individuals. Denominations, however, typically overestimate the number of members they have, and liturgical (and state-sponsored) denominations generally count anyone who has ever been baptized as a member—even infant baptisms of people who no longer claim Christian identity or attend church..
There is also another paper of the same series, Arles' World Religion Database: Realities and Concerns (2010), but I can't access its full text.
Brierley's, Woodberry's and Arles' papers were all published on the International Bulletin of Missionary Research, and Brierley, Arles and probably Woodberry as well, are/were Christian ministers and/or missionaries, and therefore I think it is important to underline that these papers belong to the Christian missionary environment to which the WCE/WRD/WCD itself belongs. Such papers are missionary sources which recommend the use of another missionary source, highlighting its strength as an unprecedented attempt to quantify the world's religious populations, while at the same time criticising its flaws. Other "high-quality publications" might be uncritical in their use of the WCE/WRD/WCD, and indeed essays like those of Brierley, Woodberry, Arles, and also Hsu et al., Stewart, and the others already discussed, were published precisely to warn against the uncritical use of such sources.
Liedhegener & Odermatt's Religious Affiliation in Europe (2013), already quoted in the 2022-2023 discussion, pointed out that (p. 9) the WCE/WRD/WCD ... is not an unproblematic source, because its data, gathered originally from the World Christian Encyclopedia, result mostly from country reports prepared by American missionaries. Therefore, a systematic bias of its data in favor of Christianity is a major, although controversial point of criticism..
As pointed out by JoelleJay hereabove, the Pew Research Center itself is very cautious in its use of WCE/WRD/WCD data, also considering that Pew mostly bases its studies on its own (real) surveys. On p. 53 of Pew's The Global Religious Landscape (2010) we read about their criteria for their use of WRD data: In cases where censuses and surveys lacked sufficient detail on minority groups, the estimates also drew on estimates provided by the World Religion Database, which takes into account other sources of information on religious affiliation, including statistical reports from religious groups themselves..
Folly Mox, in their <18:55, 5 January> comment, correctly warned that the WCE/WRD/WCD are still widely cited throughout Wikipedia in a great number of articles, mostly in infoboxes and tables and without further explanation about their nature, methodology and probable bias. This has been going on for years: many articles still uncritically report WCE/WRD/WCD data referenced to the ARDA or Gordon-Conwell websites; many of them are articles about countries and the data are reproduced directly in the country infobox, passed off as 2020 data despite the fact that they are speculative projections. Therefore, I think that it would be important that WP:CRYSTAL be mentioned in the description in the perennial sources' list.
That being said, my proposal for the description in the perennial sources' list is the following one:

There is no consensus on the reliability of these sources of data about religious populations, and concerns have been raised that they may be WP:BIASED and that they are WP:SPECULATIVE projections. Rough consensus developed that the sources should be used with in-text attribution and to prefer the use of stronger sources (e.g. censuses and national surveys). While these data sources have been used in some high-quality publications, others have questioned their methodology and consider them to be partisan, and especially prone to an overestimation of Christianity.

Æo (talk) 19:44, 6 January 2024 (UTC)

The RFC that Æo opened on this in 2022-2023 already had all of this commentary and MORE, and after all of that Firefangledfeathers was able to come up with a balanced closure wording to take into consideration ALL sides. I would say that Firefangledfeathers proposed wording, and P-Makoto's adjustment, is certainly very balanced and NPOV again and to the point. We should go with that as Firefangledfeathers is an uninvolved editor.

I also would just like to note that AEO seems to be an aggressive POV pusher against WCE/WRD/WCD sources. Seems to have an obsession to get these removed from wikipedia at any cost. To the point that he opened the 2022-2023 RFC and attempted to close it himself after the results were not in his favor (10 "No" vs 4 "Yes" - his count) with such biased wording emphasizing his view point and the minority and ignoring the majorities views (see here [11]). I thought that this opening and closing was unethical (conflict of interest) and requested an involved editor (see here [12]), which turned out to be Firefangledfeathers. His closure was much more balanced and at least took into consideration everyone's views (majority and minoirty) (see here [13]). As such, I do not trust AEO's POV pushing biased wording.

Based on this, I trust the uninvolved editor Firefangledfeathers balanced NPOV wording and P-Makoto's adjustment.

Addendum: Plus all of these quibbles were taken into account in Hsu 2008 - the only source to empirically assess these databases with 4 others: the World Values Survey (WVS), the Pew Global Attitudes Project (Pew), the Central Intelligence Agency (CIA), and the U.S. State Department (State Department) and found "The WCD is highly correlated with the other four data sets, with most correlations near 0.90, which suggests that its data for percent Christian, percent Muslim, percent Buddhist, and percent Hindu are generally reliable". Also about half or less of all countries in the world even ask about religion at all in any census. With inconsistent wording and on voluntary basis too. You have to use other sources by default to compensate. Ramos1990 (talk) 21:14, 6 January 2024 (UTC)

Actually, the only quote in my comment above that I have recovered from the 2022-2023 discussion is the quote from Liedhegener & Odermatt. Brierley's and Woodberry's texts were not quoted directly back then, except for their abstracts, and therefore my argument above provides new evidence and perspectives. Everything else in your message constitutes an ad hominem WP:PA (and I already forgave you for last year's identical one). What I have written hereabove is just my proposal building upon Firefangledfeather's one, takes into consideration all the views which have been expressed by both critical essays and editors in our discussions, and in any case I am not going to close the discussion myself. Æo (talk) 21:28, 6 January 2024 (UTC)

Addendum regarding your addendum with the quote from Hsu et al. and the consideration about census data: the WVS is not a survey specifically about religion, and it is a survey of relatively small samples (of few thousands in about 100 countries); the CIA and the US SD are not survey organisations, they collect data from some other sources (cf. Brierley himself where he states that it is not guaranteed that the CIA website is reliable); the Pew's own views are quoted in my comment above. You have to use other sources by default to compensate — yes, there is plenty of neutral statistical sources to fill gaps where we don't have data from censuses and surveys from national statistical organisations, and therefore we don't need the WCE/WRD/WCD or any other sources produced by Christian missionaries. Æo (talk) 22:04, 6 January 2024 (UTC)

I also noticed the following recently. The ARDA page for the Republic of the Congo https://www.thearda.com/world-religion/national-profiles?u=58c#IRFDEMOG has the WRD estimates as Christians making up 89.32% of the population and then breaks the Christians down as unaffiliated 9.97%, Orthodox 0.01%, Catholic 61.62%, Protestants 11.42%, and Independents 10.87%. Unfortunately adding the subdivisions up yields 93.89% which is considerably more than 89.32%. Also 89.32% fits better with the figures for other religions so it is the 93.89% that is wrong probably at least in part by overestimating the percentage of Catholics (other sources claim Catholics at 32% or 55% [taken from the State Department religious freedom report https://www.state.gov/reports/2021-report-on-international-religious-freedom/republic-of-the-congo/ which also states which 2012 government sources it got the figures from [a census and a survey]). The 2022 State department report had 47.3% Catholic and used a 2010 Pew Research Center report. The WRD database itself, which I have access to, lists 89.32% as Christian. Finding the subcategories took some work but it shows that the Christian subdivisions overlap (i.e., some people are counted in two or more Christian groups though not which groups overlap, my guess is many of those who were baptized Catholic and became something else later are counted in both which would explain why the Catholic figure is so high). However this is a guess because nowhere I can find does WRD describe their methodology (And ARDA dropped the overlap category). The list of what I assume is the sources for WRD for the Congo includes the 1960 and 2007 censuses and a 2005 survey but not apparently the 2012 government census and survey. A check on Angola also shows the double counted category missing on the ARDA listing of WRD results though it does show in the actual WRD database; however, most wiki editors do not have access to the latter. Note stuff like simple pie charts require no overlap in their data. This is even when assuming the WRD data is otherwise good data which I don't. So one can make a pie chart for the Congo using WRD data for the major religious categories (Christian, Muslim,...) but not one trying to show Catholic, Protestants, etc as well because the numbers will add up to more than 100%. Erp (talk) 01:12, 7 January 2024 (UTC)

Hey Erp. Hope you are doing well! Glad you were able to use WRD data on the Religion in Republic of Congo page. Yep, that is demography. Did you see Voas and Bruce (2004) "Research note: The 2001 census and christian identification in Britain" [14]? According to them, the British census may have overestimated Christians (71.7%) vs a common British Social Attitudes survey (54.2%). Aren't they all British who took both? Why the difference? The way a question is asked, the way a person interprets and responds play a role in differences we see in the numbers. Its more complicated with sub-divisions like denominations like "Catholic" or "Pentecostal". So I expect the variation on "Catholic" you mentioned (61.62%, 47.3%, 32%, 55%). Makes sense. With all of these numbers, it is best to let experts do the calculations than us wikieditors. They know how to use these databases better than us. In particular, sociologists of religion. Ramos1990 (talk) 03:31, 7 January 2024 (UTC)

I used it in the Congo article only because the previous editor used it and I didn't want to make too radical a change in one step. First show that the WRD data is inconsistent with itself so drop the Christian subdivisions which at least removes the inconsistencies. Then look for better sources. I'm not sure there is any really authoritative source in this case so it might be better to remove the pie chart (pie charts look nice but they lend the patina of authoritativeness which may be misleading) and discuss the different sources in the article (note some editors use multiple pie charts but that clutters up the article).

And yes how the questions are asked will affect the answers and how the survey or census is done (only resident citizens, all residents, only those with land lines...). However, the WRD isn't doing surveys or taking a census instead it is more a meta study using multiple sources (surveys, censuses, self-reported numbers, other projections) then projecting. My objections to it are several. First, it isn't clear what its sources are. The actual WRD data has a section called "Survey List" which I'm assuming is the list of sources; I have noticed in some cases that later sources than those on the list exist. Second, nothing describes the methodology it is using for a particular country; how is it calculating the projections when did it last update the projections (one can take a stab by looking at the latest item in the "Survey List" for when it likely last updated). Third there is no indication of how accurate they feel they are. Every percentage is to 4 significant digits (or counts to the individual person even when the sources aren't that precise, such as 386 people practicing Chinese Folk Religions and 237 Buddhists in the Republic of the Congo but no Daoists or Confucionists) even when that level of precision is impossible given the sources (projections should not become more precise then the sources). Another fault though common to many other sources is little account for religious syncretism such as in countries like Japan where many practice both Buddhism and Shintoism. Less important there are the oddities of definitions which make them seem not exactly neutral (for instance Confucianists have to be non-Chinese which might explain how they only get 1.8 million Confucianists in China). On another note given the use of the World Religion Database in Wikipedia for better or for worst, it is high time it had its own article complete with critiques from reliable sources so the reader can have some chance of evaluating it. Erp (talk) 16:06, 7 January 2024 (UTC)

Yeah, I don't know what the standard is in social science but in my field metastudy results should be reproducible by others, not shrouded in methodological mystery. That's another big knock on the WRD. JoelleJay (talk) 18:55, 7 January 2024 (UTC)

@Erp,

The problems you have encountered regarding the "double counting" and "inconsistent estimates" in WCE/WRD/WCD data are addressed in some of the papers we have discussed. For instance, in Hsu et al.: p. 688, analysing WCE/WRD/WCD data about US Christians: The WCD reports the total adherent count within Christian denominations and movements is 226 million, of whom 20 million are estimated to be doubly affiliated, leaving 206 million unique adherents. An additional 46 million claim to be Christians but are not affiliated with a church, for a total of 252 million affiliated and unaffiliated Christians. The 2005 Yearbook of American and Canadian Churches’ tabulation of official church membership is 163 million. In contrast to the WCD, the Yearbook does not count members of independent churches or adjust for doubly affiliated adherents. This difference of 43–63 million adherents between the Yearbook and the WCD warrants further examination. ... The WCD adjusts for “doubly counted” adherents, who may be on multiple membership lists, when aggregating up from denomination level statistics to religious blocks and total religious adherents. However, we do not know how the WCD derives its estimate of 20 million doubly counted U.S. adherents. Current WCD estimates of American Christian populations are generally higher than those based on survey evidence and denominational statistics. The WCD estimate of the total Christian population does not sufficiently reflect the recent downward trend in the percentage of Americans professing Christian identity in surveys.; pp. 689-691, analysing inconsistent estimates of Christians in other countries: We find two major groups of countries with inconsistent estimates: African countries with religious syncretism or a history of social disorder, and formerly Communist countries. ... African countries with very inconsistent estimates for percent Christian (Angola, Burundi, Congo-Brazzaville, Cote d’Ivoire, Gabon, Ghana, Zambia, and Zimbabwe) have some populations that mix religious practices. ... For India, which others have cited as problematic, the WCD has a higher estimate for percent Christian than the other data sets ... the difference comes from Christian believers in high and low castes identifying themselves as Hindu for various reasons, ... and the existence of “isolated radio believers” who do not affiliate with particular denominations. The WCE does not explain how it estimates the number of isolated radio believers, presumably a particularly difficult population to measure.. Æo (talk) 14:56, 7 January 2024 (UTC)

Further criticism of the WRD is expressed in the following paper:

Christopher Claassen & Richard Traunmüller's "Improving and Validating Survey Estimates of Religious Demography Using Bayesian Multilevel Models and Poststratification", Sociological Methods & Research, XX(X): 1-34 (Sage, 2018). On p. 4, we read: A number of data collection projects have arisen to meet this demand, including the World Religion Database ... Although the scope and comprehensiveness of these databases are admirable, and while they provide perhaps the only source of data for some regions and periods of time, there are nevertheless a number of limitations with their estimates. ... Although these databases rightly respect the adage that some data are preferable to none at all, we have no way of ascertaining the degree of uncertainty attached to any particular estimate because none are provided. Without uncertainty estimates, analysts are led to treat census measures and expert opinions as equally valid. Second, the methods used to adjust sample survey data, combine data, and obtain estimates when no data are available are less than fully transparent. Adjusting, combining, interpolating, and extrapolating data require modeling. Yet neither the assumptions underlying the model nor the exact methods for doing so are fully specified. In addition, the uncertainty induced by modeling is again ignored.. --Æo (talk) 21:12, 12 January 2024 (UTC)

RfC: WCE-WRD/WCD

The following discussion is an archived record of a request for comment. Please do not modify it. No further edits should be made to this discussion. A summary of the conclusions reached follows.

There is consensus that additional considerations should apply to the World Christian Encyclopedia, World Christian Database, and World Religion Database. There was a consensus that although scholars have advanced strong methodological critiques of the WCE and the WCD, the sources may be used with caution. Specifically, editors should attribute factual information derived from the sources and they should generally not use the sources if other reliable sources are available.

There was a strong argument that the sources should be assessed as generally unreliable based on the methodological critiques. Birdsall and Beaman 2020 was cited for a general concern around carefully defining religion, religiosity, restrictions on religion, and religious hostility in conducting surveys on religious affiliation. McKinnon 2020 was cited for its finding that WCD significantly overestimated the percentage of Anglicans in Nigeria over other surveys. Stewart 2023 was generally cited in the discussion by one editor without analysis, and it is not the closer's role to parse the source and evaluate the strength of its argument.

Despite general agreement that these concerns are legitimate, there was consensus that the sources are used with caution by reliable sources, including the Pew Research Center, Oxford Handbooks, and Cambridge reference works (some postdating the methodological critiques), and that they are published by reliable publishers—Oxford University Press, Edinburgh University Press, and Brill. There was no effective counterargument to the point that Wikipedia should do what other reliable sources do: proceed with caution. voorts (talk/contributions) 04:00, 17 February 2024 (UTC)

If you came here because someone asked you to, or you read a message on another website, please note that this is not a majority vote, but instead a discussion among Wikipedia contributors. Wikipedia has policies and guidelines regarding the encyclopedia's content, and consensus (agreement) is gauged based on the merits of the arguments, not by counting votes.

However, you are invited to participate and your opinion is welcome. Remember to assume good faith on the part of others and to sign your posts on this page by adding ~~~~ at the end.

Note: Comments may be tagged as follows: suspected single-purpose accounts: {{subst:spa|username}}; suspected canvassed users: {{subst:canvassed|username}}; accounts blocked for sockpuppetry: {{subst:csm|username}} or {{subst:csp|username}}.

The #World Christian Encyclopedia and World Religion Database/World Christian Database are currently used in many Wikipedia articles (cf. 1, 2, 3, 4) to cite statistics on religion demography, and finding a consensus on the reliability of these sources in the discussion above has been difficult. Foregoing discussions on the same sources include one in 2018 and one in 2022-2023 (with RfC).

In this request for comment, it is possible to:

Option 0: Express a comment without choosing any specific option on the reliability of the sources.
Assess the reliability of the sources by choosing from one of the following options:
- Option 1: Generally reliable
- Option 2: Additional considerations
- Option 3: Generally unreliable
- Option 4: Deprecate

Æo (talk) 18:06, 10 January 2024 (UTC)

Option 3: There is abundant evidence, especially since the publication of Stewart's 2023 critical essay, but also in previous critical essays, that these are problematic, biased sources originating from a Christian missionary environment, and they have been questioned on methodological grounds. Moreover, the data they produce are based on speculative projection. Secondary sources that recommend their use often come from the same environment, and these secondary sources express some negative criticism themselves. Secondary sources that actually use them tend to be either outdated or uncritical in ther use, often merely citing them in footnotes and/or in lists of multiple sources.--Æo (talk) 18:21, 10 January 2024 (UTC)
Bad RfC. Even if somewhere there's an unresolved discussion of the use of a cite, that still would not justify a 4-way template with options including a blanket ban. Peter Gulutzan (talk) 18:44, 10 January 2024 (UTC)
@Peter Gulutzan: What do you suggest as an alternative to the four options? Æo (talk) 18:54, 10 January 2024 (UTC)
"[Name of source] has been used in [reference to Wikipedia article] for a cite of [fact], and attempts to resolve on [name of talk page thread] have failed, please comment here." Peter Gulutzan (talk) 19:06, 10 January 2024 (UTC)
Well, both are possible. I am going to integrate the two formats. I think it is important to clearly assess the reliability of these sources, and in any case, as the rule says, "consensus is assessed based on the weight of policy-based arguments". These are not votes and the closer will judge based on the merit of all the comments here, in the discussion above, and in the previous threads as well. Æo (talk) 19:37, 10 January 2024 (UTC)

This is the standard format of RFCs on this noticeboard. The inclusion of an option shouldn't be seen as any kind of endorsement for that option. -- LCU ActivelyDisinterested «@» °∆t° 21:57, 12 January 2024 (UTC)

I doubt that anyone has authority to declare what is "the standard". Peter Gulutzan (talk) 16:44, 20 January 2024 (UTC)

Option 2. Regarding World Christian Encyclopedia (1st and 2nd ed., Oxford University Press, 1982, 2001; 3rd ed., Edinburgh University Press, 2019), World Christian Database (Brill, 2007, updated quarterly), and World Religion Database (Brill, 2007, updated quarterly): When I first saw it brought up on this board, I was inclined to encourage reassessing the sources as "Generally reliable". They are published with highly reputable university presses and have been improved across multiple editions and updates. However, after seeing the conversation between Ramos1990 and Æo, I concluded that these sources' current assessment as "Additional considerations" is fairest. Scholarly assessments of the sources evidently vary, with different perspectives about the extent to which the estimations and assessments can be depended on. As such, it makes sense to attribute these sources' projections and surveys and to be mindful of countervailing sources. However, I am not persuaded these sources should be considered "Generally unreliable". I recognize that Æo in their characterizes these sources as "originating from a Christian missionary environment". From what I have seen, that understates how the sources have emerged from an academic religious studies environment. Oxford University Press, Edinburgh University Press, and Brill don't publish just anybody, and that the editors, authors, and demographers involved met those academic standards remains meaningful. P-Makoto (she/her) (talk) 20:18, 10 January 2024 (UTC)
Option 1, Bad RFC, WP:FORUMSHOP These sources were assessed last year with the majority not supporting the same RFC poster Æo using the same arguments. I believe this may be WP:FORUMSHOP. He even tried to close the RFC himself with his own views highlighted over the majority.
In any case, these databases come from academic publishers (Oxford University Press, Edinburgh University Press, Brill). And have been independently empirically assessed too, taking into account any criticisms, with 4 other common databases in demography (the World Values Survey (WVS), the Pew Global Attitudes Project (Pew), the Central Intelligence Agency (CIA), and the U.S. State Department (State Department)) and found to be generally reliable and highly correlated (very comparable) with correlation of .9 (note: a correlation of 1 would mean perfect correlation which never happens among demographic datasets) Becky Hsu et al :"We ran correlations of the five data sets with each other on the percentage of adherents to the major world religions (Christianity, Islam, Buddhism, and Hinduism) as well as the nonreligious (Table 2). The WCD is highly correlated with the other four data sets, with most correlations near 0.90, which suggests that its data for percent Christian, percent Muslim, percent Buddhist, and percent Hindu are generally reliable". Furthermore, they note that "on the whole we find that WCD estimates are generally consistent with other data sets. The WCD is highly correlated with the other data sets, estimates for percent Christian, Muslim, Buddhist, and Hindu" and even give a positive overall recommendation "In sum, we find that the WCD religious composition data are highly correlated with other sources that offer cross-national religious composition estimates. For cross-national studies, the WCD may be more useful than other sources of data because of the inclusion of the largest number of countries, different time periods, and information on all, even small, religious groups.”
No sources have been presented showing the opposite on such a multiple global datasets scale. And the WRD methodology is available: "fully transparent to the scholarly community...based on best social science and demographic practices." It has census, surveys, polls too.
Furthermore, these sources are notable for their data being commonly commonly used by high-quality publications. They are respected by a diversity of scholars and authoritative sources such as scholars of Islam (e.g. The Oxford Handbook of Politics in Muslim Societies (2022)), scholars of nonreligion / irreligion (e.g. Annual Review of the Sociology of Religion: Volume 7 (2016)), The Cambridge Companion to Atheism (2007)), Pew Research Center's uses it in own methodology and database (see Pew's methodology, The Palgrave Handbook of Islam in Africa (2020)) and also Oxford Handbook of Mission Studies (2022), Oxford Handbook of the Economics of Religion (2011), The Oxford Handbook of Religion and Europe (2022).
If it is good enough for independent demographers, Oxford University handbooks, Cambridge University handbooks, Palgrave Handbooks, Pew, Sociology of Religion, it certainly good enough for Wikipedia. Ramos1990 (talk) 21:58, 10 January 2024 (UTC)

I respond to Ramos' argument, which I find to be misleading and which once again relies upon personal attack, in the #Additional commentary below. Æo (talk) 20:53, 12 January 2024 (UTC)

Option 1 I was originally between option 1 and 2, but after looking at some points further, I looked around and found additional high quality sources that use WRD/WCDWCE data without any issues. For example, Oxford Handbook of Religion and Europe [15] pages 793–798 uses the databases to summarize European demographics overall. I also found the same thing for summarizing demographics of Asia overall in Oxford Handbook of Christianity in Asia [16] at the very end pages 598–605. Two continents is quite good from my simple search. I think that the Becky Hsu paper on comparing WRD/WCD/WCE with 4 other secular databases with global statistics provides as good test of reliability for any given source. I was surprised such an empirical test was even done for any database at all. A .9 correlation is like an A grade for a student in school. That sociologist Robert Woodberry acknowledged Hsu's general conclusion of high correlation is a good second opinion by an expert, which is as good as it gets for global demography because demography is full of imperfections. I see no good reason for not seeing them as generally reliable at this point. Oxford, Cambridge, Pew and other unquestioned sources don't seem to either. I will lean on their expertise. After all, if WRD/WCD/WCE were unreliable, they would not even be used by them (Oxford handbooks are "Authoritative and state-of-the-art surveys of current thinking and research, from leading international figures in the discipline." [17]) The few clear criticisms I saw were minor and not significant enough compared to the positives and they were mostly Wikipedian opinions, not scholarly assessments. Pentecostalism is an informal denomination and it is hard to even get clear numbers for denominations across countries. Phil Zuckerman's struggle with atheism shows that censuses and surveys, may not be able to capture all religious groups evenly across countries and so any complaints about WRD/WCD/WCE seem to just be problems faced by demography in general, and not unique to WRD/WCD/WCE. I find it odd that the same editor opening this RFC is the same editor that opened the 2022 RFC with seeming repeat intent to depreciate again (WP:RSP entry). desmay (talk) 01:32, 11 January 2024 (UTC)
Option 3 I'm mostly going on methodological grounds and also because in my last deep dive (from the previous discussion) I found very few peer reviewed articles using the WRD as a source that were not connected with the project itself (e.g., authored by someone in WRD). Even Pew uses it only when no other sources exist. Where we know it does correlate, other better sources exist which they are probably using. One article mostly on Pew though it also applies to WRD (Birdsall, Judd; Beaman, Lori (2020-07-02). "Faith in Numbers: Can we Trust Quantitative Data on Religious Affiliation and Religious Freedom?" (PDF). The Review of Faith & International Affairs. 18 (3): 60–68. doi:10.1080/15570274.2020.1795401. ISSN 1557-0274. Retrieved 2024-01-12. notes that "Numbers are not neutral. Behind any quantification of religion or FoRB there are a range of qualitative assumptions and decisions as to what constitutes religion, religiosity, a restriction on religious belief or practice, or a social hostility involving religion. It’s both an art and a science" and goes on to state "Pay close attention to what an organization is actually measuring and use the correct terminology when citing its data. As we have seen, religious “identification” is not synonymous with faith, belief, practice, or even formal affiliation" (page 6 of the pdf). Pew almost always gives us the methodology for their figures; WRD just presents the data but not what type of religiosity they are estimating (formal affiliation, self-identification, practicing). We should also be upfront that in some cases precise numbers just aren't there so, for instance, not use a pie chart which privileges one source well above others when no source is great. (As an aside I just looked at the WRD info on the United Kingdom, I suspect it would come to a shock to many in Scotland and Northern Ireland that the UK's state religion is Anglican [it is the state religion only of England].)

Erp (talk) 02:48, 12 January 2024 (UTC)

Deep dive. Looking at an article that is comparing the WRD/WCD and several other sources (note WRD and WCD overlap on who is running them, in particular Todd Johnson) (McKinnon, Andrew (2020). "Demography of Anglicans in Sub-Saharan Africa: Estimating the Population of Anglicans in Kenya, Nigeria, South Africa, Tanzania and Uganda". Journal of Anglican Studies. 18 (1): 42–60. doi:10.1017/S1740355320000170. ISSN 1740-3553. Retrieved 2024-01-14., btw the article goes into depth about why the numbers can vary including why censuses and surveys can vary). Broadly they match other info until they hit Nigeria. "Relative to any of the other cases we have considered here, WCD estimates differ most dramatically from any of the four surveys in terms of the proportion of Anglicans in Nigeria. The WCD estimates a dramatic proportional increase in Anglicans in the 45 years leading up to 2015, from 5.2 per cent to 12.1 per cent. The highest proportion of Anglicans on any of the surveys is found in the R5 Afrobarometer survey, where Anglicans comprise 5.3 per cent of a nationally representative sample". The author continued "The WCD has arrived at its estimate for the proportion of Anglicans in 2015 by taking the last reported figures provided by the Church of Nigeria (Anglican) itself to the WCD.... In correspondence with the author, Todd Johnson of WCD has noted that, collectively, the churches and denominations of Nigeria claim 25 million more members than the best estimate of the Christian population would allow". After evaluating all the information the author concludes that there is at least 4.94 million self-identified Anglicans in Nigeria and no more than 11.74 million (the Church of Nigeria claims 18 million). BTW the ARDA report of the WRD figures (https://www.thearda.com/world-religion/national-profiles?u=166c#RELADH) which most Wiki editors likely use does not include or mention the subtraction (under the guise of multiple affiliations) that WRD uses in its own database to make the various percentages add up; the WRD total percentage of Christians is 46.18% but adding up the WRD subtotals as reported by ARDA yields 56.29%. Erp (talk) 05:11, 14 January 2024 (UTC)

Option 1 per Ramos1990 who makes a compelling argument. Nemov (talk) 14:54, 12 January 2024 (UTC)
Option 3 or at least 2. Like Erp, I am concerned with the methodology used in compiling these databases, particularly the opacity in what questions are even being asked of respondents. Pew uses WCD and other databases for only 5% of the population. That that 5% is divided into a larger number of countries than the percentages allocated to surveys etc. is about as meaningful as the observation that Trump won 2,497 counties while Biden "only" won 477. The only utility would be when discussing religious representation in the particular 57 countries that Pew used "a database" for, but in those cases we have a better source in Pew itself, which has secondarily filtered and interpreted these data. Perhaps professional demographers can extract the substantive information from WRD, but given how uneven it is in reliability and all the special considerations that one must make for given groups, we should treat it as essentially a primary source. JoelleJay (talk) 04:24, 13 January 2024 (UTC)
Option 1: Demography of religion is very sophisticated and imperfect. If it were a simple matter, or if there was a one-source solution, there wouldn't be as much debate among sociologists/demographers on religion (adherents and belief are not the same and hard to capture), and certainly not the thousands of sources that Pew needs to use to estimate religion in 232 nations. The sources in the RFC are used in quite a good number of tertiary sources from experts and reliable publishers of high-quality resources. For Wikipedia's purposes, we go by what reliable sources like these use, not if a source can solve what so many sociologists/demographers of religion have been unable to solve and continue to debate about, the number of adherents. Sociology of religion is full of debates on the estimated numbers of religious adherents. ---1990'sguy (talk) 02:38, 15 January 2024 (UTC)
Option 2: I'm fine with keeping it at option 2 but being more explicit about how it can be used. The methodology is really unclear, it is also unclear whether some numbers are predictions for the future. That the data is a lot of the time cited to ARDA instead of the real source doesn't help. WCD should never replace census data, reliable polling data or reliable membership statistics (e.g. Germany). It should not be used in religion/country articles that have better sources, the approach Pew takes essentially. If we add those other sources then I think it would violate WP:NOTSTATS and WP:UNDUE to also add the WCD. In religion/country articles where there are no better sources, it can be used but only with attribution and a disclaimer that they generally overestimate the number of christians. In the best case scenario, also an explanation of their methodology (this will be difficult). If it is used in infoboxes there should be a note with the same information. The data should not be put in a pie chart because people are far more likely to mistake it for census data or polling data (with conventional 3% error margins). --2A02:1810:BC3A:D800:A050:6C5A:A34E:91A2 (talk) 11:23, 16 January 2024 (UTC)
I'm surprised by some comments here. For instance, "The sources in the RFC are used in quite a good number of tertiary sources from experts and reliable publishers of high-quality resources"--well, that may be so, but if the argument against is "a lot of academic studies are highly critical of the source and the organization funding them is biased" and the argument for is "there are scholars that use it", then these two are not on equal footing, and it seems obvious to me that we should value those that actually studied this and other databases higher than those that simply use it/them. No, I don't find Ramos's argument compelling here, because (besides all the other problems) they are simply explaining who uses it--"If it is good enough for independent demographers..." But that, while not invalid, is simply not as strong as the counterargument, and I think the last thing we should do here on Wikipedia is use data that is published by biased organizations and questioned by scholars. So I'm actually going to go with 4: deprecate, because of its problematic source, rather than 3. Drmies (talk) 16:55, 16 January 2024 (UTC)
A request for the closer: there have been some instances of at least borderline WP:CANVASSING and there are allegations that more may have occurred, potentially off-wiki. I'd recommend adhering especially to WP:NOTAVOTE when assessing consensus. Firefangledfeathers (talk / contribs) 21:14, 19 January 2024 (UTC)
Option 2: It may be usable depending on context. Path2space (talk) 18:52, 20 January 2024 (UTC)
Option 1 per Ramos1990 and Desmay. It's trusted by extremely high-quality sources, and that means it's good enough to be used on Wikipedia. - GretLomborg (talk) 07:09, 30 January 2024 (UTC)

Additional commentary

Ramos' argument is opened by an ad hominem WP:PA (in which she falsely accuses me of WP:FORUMSHOPPING, manipulates the facts of the 2022-2023 discussion as she already tried to do in January 2023 and on 6 January 2024 in the discussion above, and accuses me of using the same arguments whilst I have presented plenty of new evidence, starting from Stewart's 2023 essay), which would be enough to make her argument fallacious. Then, she builds upon a few lines, already reiterated again and again in the discussion above, excerpted from the 2008 Hsu et al. paper which, however, is overall mostly critical of the source under discussion. Regarding the CIA and the US SD, they are not statistical institutes, and they collect statistics about religions from other sources, often from the WCE/WRD/WCD itself (e.g. US SD 2022 India report)! The Pew's very restrictive criteria in its use of WCE/WRD/WCD data have been thoroughly explained by JoelleJay and by myself in the discussion above, and once again by Erp in her comment hereabove. Then, Ramos continues by stating that no sources have been presented showing the opposite on such a multiple global datasets scale, which is misleading: various scholarly sources presented (even Hsu et al. itself!) found a systematic overestimation of Christianity and underestimation of other categories in WCE/WRD/WCD data, and various other problems, but Ramos chooses to completely ignore all the critical problems highlighted by such scholarly sources. Anne-Marie Kool's Revisiting Mission in, to and from Europe through Contemporary Image Formation (2016), another essay which is highly critical of the source under discussion, already quoted in the 2022-2023 discussion, warns that: widespread caution is raised with regard to the accuracy of the figures and not to engage in statistical analysis with the data. Æo (talk) 20:51, 12 January 2024 (UTC)

Questions of forum shopping or any other editor behaviour should be taken elsewhere, equally editors comments should be centered on sources not each other. -- LCU ActivelyDisinterested «@» °∆t° 21:59, 12 January 2024 (UTC)

Not sure about Kool. She is not a demographer and she even says on the sources, "I have taken them for as authoritative in my teaching and research during the last two decades." The Becky Hsu et al. source was not a source I found. It was Æo that cited it in the 2022 discussion and tired to use it as a main source against reliability. But after I read it, I noticed that it said the opposite and even included Christian data in the list of generally reliable data (see my blue text for quote). She is explicit on this. Furthermore, Table 2 has correlations on Christians among the 4 data sets and WCD correlates with the 4 datasets better (.9188, .9251, .9581, .9346 - all above .9 correlation) than how the other 4 datasets correlate with each other (.9146, .8979, .9365, .8468, .8538, .9408 - some are below .9 correlation). On overestimating, it is not unique [18]. Plus I found another authoritative source explicitly saying "A scholarly analysis of the World Christian Database was conducted by sociologists at Princeton University in 2008, confirming its reliability. See Hsu et al., 2008." (Bloomsbury Handbook to Study Christians (2019)) and in p. 23 acknowledges that these are "the best scholarly resource we have for documenting religious affiliation in the world today". For Pew, see my responses above. WRD is the second most used primary source after censuses, by country. Population size wise, WRD usage was comparable to large scale demographic surveys (12% of the population) - but Pew used 2,500 sources overall so it was never one source per country. Seeing that China and India alone account for 38% of the world population and all of Europe is only 7% of the global population, objections based on population size are not convincing not carry any weight. Pew goes by # of countries instead. See Palgrave handbook link for more info. Pew would simply not use WRD if it was so unreliable. Period. Numerous other authoritative sources that are commissioned specifically to leading experts in their fields (Oxford handbooks, Cambridge handbooks, etc) easily use these. Net positive, all things considered. Hope this helps. Ramos1990 (talk) 01:21, 13 January 2024 (UTC)

You continue to manipulate facts, potentially misleading readers and commentators, probably expecting them not to read the essays and the past discussions. The full paragraph of Kool (2016) in which the line you excerpted is contained is the following (underlined: your excerpt): It might well be that the great quantity of details easily silenced possible critical voices. It is peculiar that hardly any serious critical interaction and discussion of the underlying methodology of the Atlas has taken place, neither of its two data providing predecessors (footnote 65: Except for a not very convincing study: BECKY HSU et al.). The data are simply taken for granted, as I have taken them for as authoritative in my teaching and research during the last two decades.. It is a statement of repentance for having used the highly problematic WCE/WRD/WCD data in her past works.

Similarly, Hsu et al. (2008) itself (n.b. my links are always to the full paper, while Ramos' ones are always to the paper's abstract only), from which you continue to quote a few selected and decontextualised lines, is actually very critical of the sources under discussion, and I provided relevant quotes from it in the 2022-2023 discussion and others in the discussion above. It is also true that Hsu et al. is from the mid 2000s, and age matters in this case (as Doug Weller correctly pointed out in the discussion above), and therefore the excerpt you keep quoting about "high correlation" may have been true for the data of the 2000s, but no longer be true for the data of the 2010s and 2020s.

The full paragraph of Hsu et al. from which your excerpt is taken is the following (underlined: your excerpt; highlighted: critical parts): We ran correlations of the five data sets with each other on the percentage of adherents to the major world religions (Christianity, Islam, Buddhism, and Hinduism) as well as the nonreligious (Table 2). The WCD is highly correlated with the other four data sets, with most correlations near 0.90, which suggests that its data for percent Christian, percent Muslim, percent Buddhist, and percent Hindu are generally reliable. However, the other data sets often do not have information for all countries, so the correlations only represent the countries where other data sets record percentages for those religious categories. Most notably, the nonreligious data are not highly correlated between most of the data sets. While all of the data sets have mostly complete data for percent Christian and percent Muslim, data on percent Buddhist, percent Hindu, and percent nonreligious are incomplete in various data sets. The nonreligious category has few observations in State Department and CIA data and is best represented in the WCD, WVS, and Pew. The estimates for Hindus and Buddhists are especially problematic in the CIA data. Figure 1 shows that the WCD tends to overestimate percent Christian relative to the other data sets. Scatterplots show that the majority of the points lie above the y - x line, indicating the WCD estimate for percent Christian within countries is generally higher than the other estimates. Although the bias is slight, it is consistent, and consequently, the WCD estimates a higher ratio of Christians in the world. This suggests that while the percentage Christian estimates are closely related among the data sets, the tendency is for them to be slightly higher in the WCD.

Regarding the handbooks that you keep citing, they are not written by statisticians and demographers and are not essays about statistics/demography and its methodologies. They are just "handbooks" that uncritically use the WCE/WRD/WCD among many other sources. Æo (talk) 19:38, 13 January 2024 (UTC)

Key word in Hsu - "Slight". Nowhere does she say significant, nor does she isolate Christians away from the list of "generally reliable". Table 2 shows WCD had higher correlation (greater than .9) with all 4 datasets than the other 4 datasets with each other (some were below .9) on Christians as well. High correlation verifies general reliability. Overestimates/underestimates occur all the time in demography because all sources are limited. Example on census overestimating Christians too [19] and also some censuses like Soviet or Albanian censuses underestimated Christians. If WRD was as unreliable as you keep saying, high quality publications obviously would not use them even on Christianity at all, and yet they do. Among other recent ones I cited above (in my vote), here is one someone else found on summarizing Christianity in Asia [20] (btw Asia is ~60% of global population). These publications use experts in demography. Neither you or I are experts. I will leave it here. Ramos1990 (talk) 20:23, 13 January 2024 (UTC)

Hsu, whom in any case wrote although the bias is slight, it is consistent, and consequently, the WCD estimates a higher ratio of Christians, is not the only one to have found such an overestimation; almost all the other papers cited have highlighted it. For instance, Liedhegener & Odermatt found a systematic bias of its data in favor of Christianity. It is "systematic" and "consistent" throughout all countries, which means that even if the percentage of overestimation for each country were low (e.g. 3%), the overall result on the world population would be significant. The evidence suggests that in some cases the percentage of overestimation is very high: e.g. Australia WRD 2020 ~57%, cfr. Australia Census 2021 ~44% — overestimation of 13%; Canada WRD 2020 ~63%, cfr. Canada Census 2021 ~53% — overestimation of 10%; Czechia WRD 2020 ~35%, cfr. Czechia Census 2021 ~12% — overestimation of 23%; Hungary WRD 2020 ~87%, cfr. Hungary Census 2021 ~42% — overestimation of 45%; Isle of Man WRD 2020 ~84%, cfr. Isle of Man Census 2021 ~55% — overestimation of 29%; and there are many other examples. Æo (talk) 23:00, 14 January 2024 (UTC)

Your analysis: WRD vs census, is not reliable or convincing. Liedhegener & Odermatt themselves use at least a dozen sources from different time periods and adjust censuses to get their numbers for Europe. They use aggregate analysis, not single source basis (i.e. they do not just use census and that's it). Plus they admit that the quality of census can be problematic and are variable. "Even recent censuses pose sometimes serious, probably unsolvable problems to statistics on religious affiliation at a subnational or regional level. Micro censuses especially share to a certain extent the problems of survey research because the number of respondents is higher, but still restricted. The latest Swiss census is an example of the limitations to producing reliable regional, not to mention local, statistics on religious affiliation. But even traditional population censuses may cause problems which also affect regional comparisons. The British census of 2001 may illustrate this. Its results on religious affiliation where not only restricted by a missing distinction between the major Christian traditions. Moreover, the questionnaire differed substantially between England and Wales, Scotland and Northern Ireland." Some census are good for sure, but obviously not all the time. Plus less than half the countries around the world even ask about religion in their censuses. And the ones that do are inconsistently worded country by country. That means half the world is missing such data by default. Liedhegener & Odermatt themselves admit "Moreover, for Europe as a whole, as for important European countries such as France and Great Britain, it is currently impossible to give reliable figures on the religious affiliation of its population." Britain has census data on religion by the way, so what happened there is interesting. Also they do say that "For Europe the SMRE data show that the WCD provides plausible data for a number of countries, but not for all." Definitely different than your analysis above. This is an interesting admission and find it interesting that they incorporated WCD to their SMRE, it means it is indeed a valuable demographical database for SMRE. And OMG they even acknowledge historical significance as former "international “gold standard” of comparative statistics on religious affiliation, the World Christian Database". Understandable with other global datasets available than in the past. Anyways, this is not unique, "To illustrate this: The two well-known international surveys EVS and ESS use a two-stage process of questioning. In comparison to other sources this technique leads frequently to much higher results on the proportion of persons with no religious affiliation. In addition, due to different wording, the ESS produces even higher figures in this category than the EVS." Each source is limited obviously. They even say "Statistics on religious affiliation in France are a prominent example. It depends on the data you choose to either name France a catholic country or to declare it to be a highly secularized nation. However, comparing the different data on France collected by the SMRE, it becomes clear that it is virtually impossible to come up with reliable figures for this Western European country today." I like their admission that secularized countries "are countries with a lack of data or with contradicting data".

Religious demography is just a mess and not as simple as you make it seem. That is why I keep on saying that experts should be doing this stiff, not wikieditors. I rest on the experts from Oxford, Pew, Cambridge, etc on what sources are used and acceptable. Ramos1990 (talk) 00:31, 15 January 2024 (UTC)

Sorry if I'm wrong to post this as a reply.

Comment: The World Religion Database has recently (on 5 Feb.) changed it lay-out. More importantly, they have published on their FAQ a document containing their methodology. It is probably required reading for everyone giving their opinion on this source. 2A02:1810:BC3A:D800:2C98:387B:A549:4647 (talk) 19:57, 12 February 2024 (UTC)

Postponing the archiving. Will ask for a WP:CR soon. Æo, when is soon? I struggle to see what two more weeks will accomplish that the last month and a half hasn't. P-Makoto (she/her) (talk) 03:49, 14 February 2024 (UTC)

P-Makoto, Agreed. I have requested closure for both RFCs from an univolved editor to ensure fairness. Ramos1990 (talk) 00:30, 16 February 2024 (UTC)

The document has been around since 2013. Well they seem possibly to have learned about significant digits though I haven't found under the new layout what sources they are using for specific countries so that is a possible step down (even if the previous layout was unclear whether they were sources). BTW is there anyway of notifying people who have been involved in this discussion that it has new comments? I keep a watch on this page but with so many discussions going on it is hard to keep track of one. Erp (talk) 13:34, 16 February 2024 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Prigozhin vs Dr. Gilbert W. Merkx casualty numbers reliability for the infobox

There is a discussion ongoing at Talk:Battle of Bakhmut#Ukrainian losses "per Wagner" where @Alexiscoutinho insists Prigozhin casualty numbers should be kept at the infobox [21] while numbers published in Journal of Advanced Military Studies (usmcu.edu) article Project MUSE - Russia's War in Ukraine: Two Decisive Factors (jhu.edu) are not reliable enough for the infobox [22] .

My point is There are no proofs in this discussion for Prigozhin numbers reliability, and he is unreliable thus has no place in the infobox. He is unreliable per any threshold of reliability and return of his numbers is unwarranted. Balance is about balance among reliable sources and is to be achieved using reliable sources. That has all already been discussed and answered and we are in no need to repeat.

What the opinion would be? ManyAreasExpert (talk) 18:38, 6 February 2024 (UTC)

I think it's very important to really know the context of this discussion since many of my points were already exposed in the original discussion. Besides, the writing of this title seems a bit disingenuous since it mixes up, in an unfair way, two distinct concerns I raised there. Below is a transcript of the original discussion:

Extended content

Prigozhin can be used for assessing Russian losses as he would not say a number which is more than they what lose.

Assessment of Ukrainian losses however is different and he is not a reliable source for this figure. It can be left within article body but I'm removing it from the infobox. ManyAreasExpert (talk) 11:43, 5 February 2024 (UTC)

Neither side is reliable for statements of fact regarding casualties because ultimately there is no fact around them. Only Ukraine, Russia and Wagner know their own losses and this won't come out until well after the war. Saying one estimate is more reliable than the other when they ultimately can't be seems like cherry picking to me. Furthermore, it makes the infobox imbalanced. Show both sides in a balanced way and with proper atribution, or show none. Alexis Coutinho (talk) 15:23, 5 February 2024 (UTC)

Neither side is reliable for statements of fact regarding casualties
— User:Alexiscoutinho 15:23, 5 February 2024 (UTC)

No, US estimates, BBC estimates, and Journal of Advanced Military Studies article estimates are much more reliable regarding losses, and Prigozhin has no reliability at all. Even Ukraine estimates are more reliable then Russian sources estimates.

Saying one estimate is more reliable than the other when they ultimately can't be seems like cherry picking to me
— User:Alexiscoutinho 15:23, 5 February 2024 (UTC)

No, that's stating facts. Prigozhin estimates have no reliability at all.

Furthermore, it makes the infobox imbalanced. Show both sides in a balanced way and with proper atribution, or show none.
— User:Alexiscoutinho 15:23, 5 February 2024 (UTC)

The balance is achieved among and using reliable sources. We are to not to use unreliable sources to achieve balance. ManyAreasExpert (talk) 20:01, 5 February 2024 (UTC)

Still dispute. Wagner's estimate, the main party in the battle, is at least quite relevant. This feels like an abuse of the argument of reliability, maybe a simplistic interpretation of the principle. By the way, this is all WP:AGF, that's why I'm using words like "feels like" and "maybe"... While I agree that an academic source is surely more reliable, I thought it was quite inadequate to keep the "less reliable" estimates only tending towards one side. This ties to your initial comment of: Prigozhin can be used for assessing Russian losses as he would not say a number which is more than they what lose. Assessment of Ukrainian losses however is different and he is not a reliable source for this figure., which I found quite concerning as it seems like a crude and convenient oversimplification. I would really appreciate if other uninvolved editors weighted in on this too. Alexis Coutinho (talk) 20:23, 5 February 2024 (UTC)

Still dispute. Wagner's estimate, the main party in the battle, is at least quite relevant.
— User:Alexiscoutinho 20:23, 5 February 2024 (UTC)

It's called notable, not relevant. That's why we include it into article body, attributed. That doesn't make it reliable.

it was quite inadequate to keep the "less reliable" estimates only tending towards one side. This ties to your initial comment of: Prigozhin can be used for assessing Russian losses as he would not say a number which is more than they what lose. Assessment of Ukrainian losses however is different and he is not a reliable source for this figure., which I found quite concerning as it seems like a crude and convenient oversimplification.
— User:Alexiscoutinho 20:23, 5 February 2024 (UTC)

Russian-side sources are widely unreliable and this is their problem, not wikipedia's. We are still to not to use fringe sources to achieve a false "balance". ManyAreasExpert (talk) 20:32, 5 February 2024 (UTC)

Another thing. The purpose of infoboxes is to summarize and give an overall view of the content of the article. This implies fairly and in a balanced way representing the views of the article. If in the process it makes an unbalanced representation of the content, then it's not fulfilling its purpose and is doing something wrong. It doesn't make sense for only the infobox to have a higher threshold of reliability. Alexis Coutinho (talk) 15:48, 6 February 2024 (UTC)

Greetings, there are no proofs in this discussion for Prigozhin numbers reliability, and he is unreliable thus has no place in the infobox. He is unreliable per any threshold of reliability and return of his numbers is unwarranted. Balance is about balance among reliable sources and is to be achieved using reliable sources. That has all already been discussed and answered and we are in no need to repeat. ManyAreasExpert (talk) 17:55, 6 February 2024 (UTC)

There is also no proof that those other estimates are more reliable for statements of fact, they don't refer to confirmed numbers after all. The publishing sources are all generally reliable. However, there is no proof that the figures of each viewpoint/party are really accurate. As such, we give WP:DUE weight to each viewpoint and properly attribute them. I fear that unless more editors participate here or this discussion is moved to WP:RSN, the discussion won't progress.

As an alternative, I propose moving all the varying estimates of strength and casualties from the infobox section to the body of the article and keep only a generic word like "Heavy" for each side or just link to the sections, like was done in the 2023 Ukrainian counteroffensive page. This would be the least controvertial and most impartial solution and would clean up the infobox. We should only keep facts there and information that there is consensus on. What do you think? Alexis Coutinho (talk) 18:27, 6 February 2024 (UTC)

Wikipedia:Reliable sources/Noticeboard#Prigozhin vs Dr. Gilbert W. Merkx casualty numbers reliability for the infobox ManyAreasExpert (talk) 18:39, 6 February 2024 (UTC)

Dr. Gilbert W. Merkx reliability

In fact, I would ask you to demonstrate how that "academic source" you cited is objectively "more reliable" than all the other estimates. I've checked the publication and I really wasn't impressed. Firstly, it's a review piece that talks about the war as a whole, thus talks about a bit of everything. The only part I found that was related to casualties in the battle of Bakhmut was this: In Bakhmut, the most intense of these battles, recent estimates suggest that Russian forces suffered between 32,000 and 43,000 dead and 95,000 wounded, with Ukrainian losses at about 15–20 percent of that. That's it. In a sea of other text, the author only says this. And it doesn't even seem like he made the estimate. It seems like he's drawing from other estimates. Which estimates? Ukrainian estimates? US estimates? It's not clear and maybe we just can't know. The only nearby reference [45] is from a NYT article from the start of the war. So I really don't see much credibility with that statement. At first glance, that "academic ref" just seems like an adornment and, as such, I'll partly comment it out until consensus is gained for its inclusion. Alexis Coutinho (talk) 20:41, 5 February 2024 (UTC)

The article Project MUSE - Russia's War in Ukraine: Two Decisive Factors (jhu.edu) is published in Journal of Advanced Military Studies (usmcu.edu) and JAMS uses a double anonymous peer review process to evaluate submissions. Subject matter experts who specialize in military history, national security, international relations, social science topics are recruited from internal and external agencies to support JAMS's annual publishing process. ManyAreasExpert (talk) 21:02, 5 February 2024 (UTC)

This is the author's attribution: Dr. Gilbert W. Merkx is professor emeritus, Department of Sociology, Duke University, where he also served as vice provost for international affairs and director of the Center for International Studies. He has published eight books and 200 articles and chapters, a number of which deal with military and security issues. Looks to me like he knows a bit (or enough) of many things. In fact, his degrees are all in Sociology. I'm not saying he's lying. Those estimates he's referring to are very likely real. But the thing is, it doesn't seem like those are his estimates, so his qualifications are mostly useless to that statement. That is not an article dedicated to battle of Bakhmut, let alone the casualties of it. Therefore, the citation is malformed or at least misrepresented. Furthermore, it doesn't mean very much what JAMS talks about itself. There are plenty of "bad journals" out there that would say good stuff about themselves. Ideally, another source should say how qualified JAMS is, or simply other well respected editors here. But once again, I'm not implying that JAMS is unreliable/lying, I'm willing to believe it is good even without knowing much about it. I aimed mostly to nullify that unsatisfactory argument as a sole response, ie. I wanted you to use a stronger argument to convince me or advance the discussion elsewhere (ie. how to reach a compromise instead of who's right or wrong). Alexis Coutinho (talk) 21:19, 5 February 2024 (UTC)

Oh, and another thing I forgot to mention, even if there is indeed a double anonymous peer review process, I wouldn't be surprised if not all statements were fully fact checked. I believe the purpose of the check is to mostly validate the general conclusions and most important statements of the publication. How could the reviewers fact check vague/opaque statements like that? This is not natural sciences... Alexis Coutinho (talk) 21:26, 5 February 2024 (UTC)

This journal is used more than 150 times " Journal of Advanced Military Studies" site:wikipedia.org - Google Search currently so I'd say there is a consensus regarding its reliability now. ManyAreasExpert (talk) 21:28, 5 February 2024 (UTC)

Just note that this isn't my point btw. As a side note though, reliability of a journal doesn't really come from the number of Google search hits, it comes from the amount of citations in other journals. I'm not asking you to provide this info and I get your point of reliability of JAMS. I'm just trying to help you out (make stronger arguments) for maybe future discussions. Alexis Coutinho (talk) 21:31, 5 February 2024 (UTC)

wp:rsn. Slatersteven (talk) 21:19, 5 February 2024 (UTC)

So, summarizing, there are two concerns:

All varying estimates shown in the infobox are published by reliable sources. These can be trusted to be accurately making statements of fact. What are these statements of fact? To faithfully relay the estimates from the original primary source. Now here comes the problem, all of those primary estimates are, well, estimates, they are not confirmed casualties (or confirmed strength) and all have varying degrees of uncertainty. So it seems inadequate to be extending the notion/argument of reliability to those primary sources/estimates aswell. Furthermore, those estimates all come from heavily involved parties: Ukraine, Wagner and USA (which spent a massive amount of money in the war and naturally has its own interests to justify all that past and continued spending). I believe it's too early to be making assertions that one estimate is more reliable than the other (and then trying to hide one estimate and advertise the other) when the correct terminology should actually be "how likely" each one is to be closer to the true number. And the true number is something only Russia and Ukraine know for their own troops; a figure which won't appear until well after the war. I also thought that ManyAreasExpert's argument that Wagner's estimate was unreliable for the infobox, but was acceptable for the article body was quite inadequate. There should be no varying or arbitrary thresholds of reliability in different parts of the article. The infobox should summarize the points/views presented in the article in a balanced way. If, in the process, it skews the balance by filtering out particular views/information, then it's not serving its purpose and is actually being misleading.
The other concern is about Dr. Gilbert W. Merkx's citation. While ManyAreasExpert attempted to prove the reliability of the author and journal, I pointed out that his method wasn't ideal. This/here is a better place to reach such conclusions about reliability. But even though I didn't find his proof particularly satisfying, that wasn't my point. I argued that the attribution of his citation in the article was incorrect. ManyAreasExpert wrote the citation implying that Dr. Gilbert W. Merkx made the estimate. But closer inspection of the source showed that this likely wasn't the case. The author, with degrees in Sociology but still knowledgeable in other relevant areas (I showed his full attribution in the original discussion collapsed above), in a review paper that talked about the war as a whole, made a passing mention of recent estimates. This is the quote in the publication: "In Bakhmut, the most intense of these battles, recent estimates suggest that Russian forces suffered between 32,000 and 43,000 dead and 95,000 wounded, with Ukrainian losses at about 15–20 percent of that." That's the only thing the author says in a sea of text. It seems likely that he isn't refering to his estimate (which is good since he doesn't seem to have the ideal qualifications to make one). But then whose estimate is that? American, Ukrainian, German? We don't know and maybe can't. Possibly even the peers who allegedly review JAMS pulications might not have exactly known the source. After all, the publication had a completely different objective/thesis. In light of this, I corrected the citation here and the problem should have been solved. Given that the primary source of the estimate is unknown, one must not attribute it directly to the secondary academic source. Thus I don't fully understand why ManyAreasExpert is bringing this up again. Perhaps to fully clarify the matter...? Which would be understandable.

Damn, this was supposed to be a summary, but ended up being quite long. If something I wrote isn't quite clear, then reading the original discussion will probably be more elucidating. I think the real juice is there.

Oh, and I must not forget that I suggested to move all those quite different estimates from the infobox to the body of the article. There, we can properly give WP:DUE weight to all notable views/estimates in a balanced way. It also clears up the infobox and let's us keep only the facts and uncontrovertial information: that both sides suffered "heavy" losses and that the reader should also check the appropriate section to really understand who said what and by how much they differ. Alexis Coutinho (talk) 00:03, 7 February 2024 (UTC)

Infoboxes are in general bad places to try and discribe complex disagreements. I would suggest changing them to "lowest estimate<refs> – highest estimate<refs>" and moving any furthet details into the article content. There who made estimates and any criticism of those estimates can be explained in full. I would also suggest all estimates are attributed as estimates are.. well estimates. Also attribution should be to the source, for instance the current infobox attributes a report by Sky News as "Western estimates" as if Sky News is the mouth piece of of some multi-national conglomerate. It should either be attributed to Sky News or the source that Sky News is quoting (if any). The same issue holds true for several of the other attributions (whether 'western' or not). -- LCU ActivelyDisinterested «@» °∆t° 16:12, 9 February 2024 (UTC)
Prigozhin numbers are unreliable to be included into infobox no matter if they are attributed or not.
Sky News cites Western officials and there are other media outlets for the same number so we don't have to be that precise with attribution and just settle on "Western".
Or, we can use academic article referred above. ManyAreasExpert (talk) 16:23, 9 February 2024 (UTC)
Sorry to interrupt, but your previous reply gives the impression that you're not looking for advice here. Alexis Coutinho (talk) 17:06, 9 February 2024 (UTC)

I've just re-read the Sky article again just to be sure, and it never mentions Western officials or any other source for it's figures in any way. To be clear I mean this article[23] that is used to say 'Western officials' claim Ukrainian loses are 20,000+ killed or wounded.
The sources for an estimate by a western official of Russian nukbers we're both just quoting a different Sky article, I've replaced both with the direct Sky News article.[24]
Prigozhin would be considered reliable for the losses according to the Wagner group, in the way that any group is reliable for their own words. Putting these details in the content on the article allows for any criticism of them to also be added, something the infobox can't handle. -- LCU ActivelyDisinterested «@» °∆t° 22:11, 9 February 2024 (UTC)
Thanks! Ukraine-Russia war latest: Zelenskyy replaces another top army chief; Putin wanted to show he's 'ready for peace talks' in interview | World News | Sky News says More than 60,000 Russian casualties have been reported in Moscow's efforts to capture the Ukrainian city of Bakhmut, according to a Western official and there are other sources [25] [26] . ManyAreasExpert (talk) 22:32, 9 February 2024 (UTC)
Yes as I explained that's not the part I'm talking about. I as talking about the source used in the infobox for 20,000+ killed or wounded Ukrainian loses which doesn't mention western sources. -- LCU ActivelyDisinterested «@» °∆t° 22:49, 9 February 2024 (UTC)

Prigozhin would be considered reliable for the losses according to the Wagner group, in the way that any group is reliable for their own words. Putting these details in the content on the article allows for any criticism of them to also be added, something the infobox can't handle.
— User:ActivelyDisinterested 22:11, 9 February 2024 (UTC)
... and that means ... ? Thanks! ManyAreasExpert (talk) 20:28, 10 February 2024 (UTC)
Something like "Yevgeny Prigozhin estimated the Wagner group loses to be ..." If I published an article saying I believed the worlds favourite fruit was the orange, the you could attribute as "ActivelyDisinterested says the worlds favourite fruit is the orange". Prigozhin as the leader of the Wagner group is reliable for statements from the Wagner group. You can't use that to say it's true in wikivoice, but you can use it to say that what he said. -- LCU ActivelyDisinterested «@» °∆t° 20:44, 10 February 2024 (UTC)
There is already a paragraph saying this in article body.
The argument is that why we should have Prigozhin numbers in the infobox. ManyAreasExpert (talk) 20:49, 10 February 2024 (UTC)

If an argument is being made against JAMS as a reliable source it needs to be spelled out better. I wouldn't attribute them as academic sources, as that's a very wide net and just invites {{who?}}, but "an estimate in the Journal of Advanced Military Studies" could work (or an estimate by Gilbert Merkx in the Journal etc..). -- LCU ActivelyDisinterested «@» °∆t° 22:19, 9 February 2024 (UTC)
I would admit Merkx's estimates seem high, but they are also appear newer and independent. -- LCU ActivelyDisinterested «@» °∆t° 22:21, 9 February 2024 (UTC)
newer and independent should we actually assume that? Alexis Coutinho (talk) 22:25, 9 February 2024 (UTC)
That would depend on your argument against Merkx and JAMS reliability. -- LCU ActivelyDisinterested «@» °∆t° 22:50, 9 February 2024 (UTC)
Just found this in Wikipedia:Reliable sources#Context matters which may be useful: Information provided in passing by an otherwise reliable source that is not related to the principal topics of the publication may not be reliable; editors should cite sources focused on the topic at hand where possible. Alexis Coutinho (talk) 14:09, 10 February 2024 (UTC)
All these policies have "may not" or "should' in them, instead of "will not" and "must". They are guidance and advice rather than scripture.
However saying that I spent a couple of hours trying to find any collaboration for this estimate. Merkx does say it's someone else's estimate, but doesn't say who and doesn't give a citation (oddly given everything else is properly cited in the article). The only thing I can find coming close is some old leaked estimates from US officials, but they were for the whole conflict at the point not just the battle of Bakhmut. I'm concerned that the numbers are a mistake, even in reliable sources they do happen. -- LCU ActivelyDisinterested «@» °∆t° 20:52, 10 February 2024 (UTC)

"an estimate in the Journal of Advanced Military Studies" could work
— User:ActivelyDisinterested 22:19, 9 February 2024 (UTC)
Thanks! Right now it says Fall 2023 research suggests that recent estimates[who?] have Russian losses at 32,000–43,000 dead and 95,000 wounded and Ukrainian losses at 15–20% of the Russian's in the article body. ManyAreasExpert (talk) 22:33, 9 February 2024 (UTC)
You seem to have misunderstood slightly, the "[who?]" part is highlighting a problem. This seems to be an estimate inside another work rather than research dedicated to the specific subject, it shouldn't say who made the estimate rather than "Fall 2023 research". -- LCU ActivelyDisinterested «@» °∆t° 22:45, 9 February 2024 (UTC)
"Fall 2023 research" is fine but under what attribution to include it into the infobox? Thanks! ManyAreasExpert (talk) 20:24, 10 February 2024 (UTC)
It is not fine, as it is not attribution. It's not even research but Merkx stating someone else's estimate. If it's going to be included, and I'm leaning more than it shouldn't, it should be attributes to Merkx or JAMS. -- LCU ActivelyDisinterested «@» °∆t° 21:22, 10 February 2024 (UTC)
Agree with this attribution as well. ManyAreasExpert (talk) 21:37, 10 February 2024 (UTC)

I'm going to create two new sections, as this is becoming confusing. -- LCU ActivelyDisinterested «@» °∆t° 21:35, 10 February 2024 (UTC)
I've notified the WP:MILHIST project to hopefully get some extra input to the discussion. -- LCU ActivelyDisinterested «@» °∆t° 21:46, 10 February 2024 (UTC)

Yevgeny Prigozhin and the Wagner group

Is Yevgeny Prigozhin and the Wagner group reliable for estimates of loses suffered by the Wagner group during the battle of Bakhmut? Should Wagner groups estimates of Ukrainian loses be included in the infobox? -- LCU ActivelyDisinterested «@» °∆t° 21:35, 10 February 2024 (UTC)

Per Cinderella's comment in the section below, I see it as premature to attempt to judge the reliability of estimates from either side when no good quality independent assessments currently exist. The current estimates are all flawed to a varying extent, and as such, should only be explored in the article body. Alexis Coutinho (talk) 01:52, 11 February 2024 (UTC)

I'd be very wary of using such sources in an infobox as by definition unreliable, but might be OK with a clear "per Wagner" attribution. BobFromBrockley (talk) 19:56, 12 February 2024 (UTC)

Please @Alexiscoutinho, you are again putting on the same bench an academic research and Prigozhin claims [27] . ManyAreasExpert (talk) 09:30, 16 February 2024 (UTC)

I've added intext attribution to clarify where the lower and upper bounds are coming from. The casualties summary statement should be ok now. Alexis Coutinho (talk) 16:07, 16 February 2024 (UTC)

Well, it's pretty clear that you still didn't like it. Alexis Coutinho (talk) 23:52, 16 February 2024 (UTC)

It goes without saying that each side of a conflict has many reasons not to be fully truthful about their own and the enemy's losses.

In this case the question is whether these numbers are cited by reliable sources. I see that the Jerusalem Post cited Prigozhin's estimate of the Ukrainian forces (80k) [28]. Are the losses numbers also cited by RS? Alaexis_¿question? 23:14, 16 February 2024 (UTC)

Yes. Politico and Kommersant. Alexis Coutinho (talk) 23:49, 16 February 2024 (UTC)

No, any claims of this nature by Prigozin should be regarded as a disinformation. Yes, they were cited by a variety of sources as a claim by Prigozin, but it does not mean they are not disinformation. Same applies to claims by Russian Ministry of Defense, claims by international terrorist organizations (such as Hamas), etc. Yes, their numbers may be cited (just as claims by conspiracy theorists, etc.), but we must make it clear on the page they are not to be trusted. My very best wishes (talk) 03:05, 17 February 2024 (UTC)

🤦‍♂️ No, any claims of this nature by Prigozin should be regarded as a disinformation. Your personal opinion. Also, see WP:INTEXT. Alexis Coutinho (talk) 04:27, 17 February 2024 (UTC)

Gilbert W. Merkx and the Journal of Advanced Military Studies

Are the estimates given by Merkx in this article[29] in the Journal of Advanced Military Studies reliable? -- LCU ActivelyDisinterested «@» °∆t° 21:35, 10 February 2024 (UTC)

The rest of Merkx's article is well cited but the estimates give no citation and I can't find anything to back up the estimates. The exact quote is recent estimates suggest that Russian forces suffered between 32,000 and 43,000 dead and 95,000 wounded, with Ukrainian losses at about 15–20 percent of that. These don't appear to be an Merkx estimate and the only thing close I can find is an estimate for the entire war not just the battle of Bakhmut. -- LCU ActivelyDisinterested «@» °∆t° 21:40, 10 February 2024 (UTC)

I would actually greet any other academic estimate as well, including, for example, ISW's. Couldn't find another one yet. ManyAreasExpert (talk) 21:52, 10 February 2024 (UTC)

I spent some more time looking for estimate, but estimates seem hard to come by. Estimates for Russian figures are all very similar to the figures given by Wagner, while finding anything for Ukrainian figures is proving extremely hard. -- LCU ActivelyDisinterested «@» °∆t° 22:01, 10 February 2024 (UTC)

Per WP:INFOBOXPURPOSE, the infobox is to summarise key points from the article. Placing multiple estimates in the infobox for casualties is intricate detail for which the infobox is unsuited. We do not have casualties in the infobox for Russian invasion of Ukraine, partly because the war is ongoing and partly because there are no good quality independent assessments to draw upon and report in the infobox with confidence. We should note that what is reported in an infobox is often viewed as fact. Even though the fighting for Bakhmut has concluded, one can say that the dust is yet to settle and the circumstances for Bakhmut are not unlike the ongoing war. We lack good quality independent assessments and are unlikely to until the dust has well and truly settled. We are not obliged to populate the infobox casualty section. There is WP:NODEADLINE. In the meantime, we can write in the body of the article, a summary of what has been published about casualties and the TOC will direct our readers to that section if it is their desire to know more about casualties. Cinderella157 (talk) 00:49, 11 February 2024 (UTC)
Completely agree. Alexis Coutinho (talk) 01:47, 11 February 2024 (UTC)

Having spent many hours now looking for estimates and such for the battle specifically I think this is the best option. -- LCU ActivelyDisinterested «@» °∆t° 18:58, 11 February 2024 (UTC)

Some resources that covers comics information.

Hello, my user name is Sewnbegun. I am new to Wikipedia and here for editing mainly lists/tables (obviously not exclusively) regarding comics, tv series and films. I searched many sources which I though can be reliable sources; in the list at WP:RSP. I was surprise that only three (Screen Rant, IGN and Gizmodo) are considered as reliable source, for one (Dexerto) is advised to find alternative source while remaining are missing (including the ones that constantly tells about comics - like Comic Book Resources and Aiptcomics). So I am starting a new enquiry about that remaining sources of my list and can somebody help me to know which of them are reliable?:

Comic Book Resources
AIPT
ComicBook.com
SuperHeroHype
Official website of Marvel (Marvel.com)
GamesRadar+ (for comics)
Bleeding Cool News
Popverse

Sewnbegun (talk) 18:13, 18 February 2024 (UTC)

You may find the WikiProject Comics sources and WikiProject Video games sources pages helpful.

Besides that, it's important to understand that source reliability always depends on the context. For example, Marvel.com may be a good site for primary sources like comic books or claims about its corporate structure, but we should generally cite reliable, independent, secondary sources for most content. ComicBook.com is ostensibly "independent", but its articles often include affiliate links, which may present a bias or conflict of interest in its coverage. The reliability summaries at RSP are also general and may not apply in all cases.

To answer your question, we'd really need to know the specific sources you'd be citing, and what you'd like to cite them for. Woodroar (talk) 20:15, 18 February 2024 (UTC)

Thanks @Woodroar, I got your point and now I am going cite few sources that can leads to more clearity.

Aipt - ^[1]
Bleeding Cool - ^[2]

Sewnbegun (talk) 01:43, 19 February 2024 (UTC)

The AIPT source is an interview, so anything Duggan says would fall under WP:PRIMARY and WP:ABOUTSELF, and also carry no weight as far as notability is concerned.

The Bleeding Cool source looks like a roundup of stories about X-Men cancellations from the day before. Instead, you should use and cite the original source.

These are just general tips, however. Everything depends on what you want to use the sources for. Woodroar (talk) 02:04, 19 February 2024 (UTC)

@Woodroar, what about this Bleeding Cool reference - ^[3]? Sewnbegun (talk) 02:12, 19 February 2024 (UTC)

I would think that source is fine for, say, plot highlights and credits from those two issues. Primary sources can sometimes be used for those things, but coverage in a secondary source (Bleeding Cool) gives it more editorial weight or importance. Woodroar (talk) 02:42, 19 February 2024 (UTC)

Thanks for helping! Sewnbegun (talk) 02:47, 19 February 2024 (UTC)

Glad to help! Woodroar (talk) 03:00, 19 February 2024 (UTC)

References

^ Hassan, Chris (August 14, 2023). "X-Men Monday #215 – Gerry Duggan Talks 'X-Men' at FAN EXPO Boston 2023". AIPT Comics. Retrieved February 18, 2024.
^ Johnston, Rich (February 18, 2024). "Marvel Cancels All X-Men Comics in the Daily LITG, 18th February 2024". Bleeding Cool. Retrieved February 19, 2024.
^ Johnston, Rich (February 7, 2024). "Invasion, Insurrection & Kissing in Today's X-Men Comics (XSpoilers)". Bleeding Cool. Retrieved February 19, 2024.

Al Arabiya

Is this an RS (giving its dodgy rep) especially for someone being dead? Slatersteven (talk) 12:46, 14 February 2024 (UTC)

Probably OK if nothing to do with Saudi. Selfstudier (talk) 12:53, 14 February 2024 (UTC)

Muhammad al-Zawahiri, his death (odd this is a saudi source, and not an Egyptian one). Slatersteven (talk) 12:56, 14 February 2024 (UTC)

Hum , might look for confirmation in that particular case. Selfstudier (talk) 13:05, 14 February 2024 (UTC)

Yes well the person who is adding is says it's enough. Slatersteven (talk) 13:06, 14 February 2024 (UTC)

@Slatersteven:, a quick glance at his article would have told you that he owned a construction firm in Saudi and moved there to live at some point in his life. I therefore don't find it strange that news of his death was broken by a Saudi news outlet? Somebody's birth does not determine where their obituary comes from if they relocated throughout their lifetime. --Jkaharper (talk) 17:50, 14 February 2024 (UTC)

50 years ago, and long before his arrest in Egypt. Slatersteven (talk) 17:53, 14 February 2024 (UTC)

Whats the dodgy rep for al-Arabiya? nableezy - 17:57, 14 February 2024 (UTC)

Being sanctioned by Ofcom, which led to them surrendering their UK broadcast license, for a start? Slatersteven (talk) 18:00, 14 February 2024 (UTC)

That seems to be regarding this. I dont really think a British government agency is who we should be relying on to determine the reliability of a source. nableezy - 18:04, 14 February 2024 (UTC)

Ofcom is an independent body, and it was after more complaints were made, and were about to be investigated they dropped their license. Then there is plagiarism. The use of fake reported to plant stories. Slatersteven (talk) 18:09, 14 February 2024 (UTC)

Independent of the political branches sure, but still an agent of the British state. Sources for plagiarism and planted stories? Im genuinely asking, Im not saying I know this source is reliable, I just want to see the evidence that it is not. nableezy - 18:16, 14 February 2024 (UTC)

Slatersteven, that Ofcom ruling appears to be about bias/framing on political matters, which is a different matter to generally questioning reliability. Ditto the above user's comments – Ofcom isn't an authority on Wikipedia by which we give any additional weight to because it's tied to the UK Government. We need to judge sources independently of the opinions of any other bodies or organisations. --Jkaharper (talk) 18:19, 14 February 2024 (UTC)

Err, no, as RS have to have a good reputation, in other words, other people regard them as reliable. And (again) it's not just ofcom. Slatersteven (talk) 18:21, 14 February 2024 (UTC)

I don't think it's exactly right that we judge sources independently of the opinions (or in this case findings) of other bodies. We weigh up what other sources say. An independent broadcasting regulator or press regulator, while not necessarily the final word, is a good source to take into account in assessing reliability. The Ofcom ruling was not about bias/framing, it was about use of false testimony extracted under torture, a fundamental issue of both journalistic ethics and reliability. However, a single instance isn't enough to make a general unreliability ruling - are there other instances? And does it relate to specific topic areas (e.g. Gulf geopolitics) or is it a general issue? BobFromBrockley (talk) 11:04, 19 February 2024 (UTC)

Per Wikipedia:Reliable_sources/Noticeboard/Archive_418#Al_Arabiya, it is a usable source except where the Saudi government could be reasonably construed to have a conflict of interest. Hemiauchenia (talk) 23:36, 14 February 2024 (UTC)

I'll leaved it for today, I will get back to you. Slatersteven (talk) 18:21, 14 February 2024 (UTC)

Wait for what exactly? You don't seem to be winning any support from your arguments, Slatersteven. Fact is unless you have clear grounds to dispute the reliability of a news source, then editors should be allowed to cite it. Not any agreement above nor on the edit history of the Muhammad al-Zawahiri page. His death has been up on the Deaths in 2024 page all day, and despite 1000s of active users visiting that page in that time, it hasn't been brought into question. If you can bring forward evidence we can discuss this further but in the mean time, the source seems sufficient. --Jkaharper (talk) 23:19, 14 February 2024 (UTC)

The request for sources about its faults. [[30]] [[31]] [[32]] [[33]]. Slatersteven (talk) 11:18, 15 February 2024 (UTC)

bnnbreaking.com ?

I've seen this pop up in searches now and then, aboutpage:[34]. Where does it fall on the newspaper "general reliability" scale? Gråbergs Gråa Sång (talk) 11:59, 15 February 2024 (UTC)

Unreliable

they make unverified claims on About page and say they are a source for The Gateway Pundit a website known for publishing falsehoods, hoaxes, and conspiracy theories Softlem (talk) 12:09, 15 February 2024 (UTC)

We have an article on them: BNN Breaking. It's not encouraging. From some of the sources:

One of those stories, which was bylined by BNN Breaking founder Gurbaksh Chahal and was riddled with inaccuracies; The sheer volume of stories on the site raises questions about whether they’re being generated with the help of artificial intelligence tools, according to Hany Farid, a professor at the UC Berkeley who specializes in deep-fake technology. [35]
The site BNN News made the false report...; Instead of making a correction... [36]
Twitter said it is suspending scores of new accounts purporting to be part of a "news network" called BNN [37]

Barnards.tar.gz (talk) 12:19, 15 February 2024 (UTC)

Thanks for the replies! Apparently there is a bit of a "feud" going on:[38]. FYI @Chisme and @Lepricavark, you're on bnn. Clever name, I think. Gråbergs Gråa Sång (talk) 12:49, 15 February 2024 (UTC)

Thanks for the heads' up, but I'm not gonna bother giving them a click. As for the question being addressed in this thread, I doubt that any publication founded by convicted woman beater Gurbaksh Chahal is reliable, given his extensive history of interfering with our article on him. I'm sure he has an axe to grind against me, but when you remember that he was caught on camera smothering the woman with a pillow and hitting and kicking her 117 times over a half hour period [39], perhaps he should be focused on dealing with other problems in his life. LEPRICAVARK (talk) 14:15, 15 February 2024 (UTC)

Well, that's one reason for him have his own media-org. Better press. Gråbergs Gråa Sång (talk) 15:20, 15 February 2024 (UTC)

Looking at the BNN Breaking article, having "Criticism" as the first and only prose-section is not good WP-writing. Gråbergs Gråa Sång (talk) 12:56, 15 February 2024 (UTC)

The more I look at this, the shadier it gets. The whole site appears to be a clickfarm built out of AI-laundered content taken from legitimate news sites. It's not clear that the "journalists" are even real people. This is all extremely corrosive. I suggest the site be blacklisted as spam. Barnards.tar.gz (talk) 14:42, 15 February 2024 (UTC)

The images are a dead giveaway:

Here's an article about "Brave London Police Officers", which features File:Police in Glasgow.jpg from Wikipedia Commons. (Glasgow notably not being in London). There is no attribution as required by the license.
This article uses an illustration that came from this artist. I contacted the artist who confirmed that this use appears to be straight-up copyright infringement.
This article about recycling is illustrated with a completely unrelated image of a baseball player. Surely if a human was involved in the editorial process, they would have noticed this?

It's garbage and all links to it need to be removed per WP:LINKVIO. I count 181. Barnards.tar.gz (talk) 15:21, 15 February 2024 (UTC)

You search better than I do. Gråbergs Gråa Sång (talk) 15:25, 15 February 2024 (UTC)

Interesting idea. Not much presence on WP [40], not under that name at least. Gråbergs Gråa Sång (talk) 15:23, 15 February 2024 (UTC)

User contributions for Bnnbreaking Gråbergs Gråa Sång (talk) 12:52, 16 February 2024 (UTC)

Blacklist This is a worthless content farm that has no place on Wikipedia. Hemiauchenia (talk) 22:02, 15 February 2024 (UTC)
This is not an RS, but it's in-depth: The Curious Case of BNN Gråbergs Gråa Sång (talk) 08:18, 16 February 2024 (UTC)

I believe WP:RSN is appropriate not for questions about "general reliability" but for specific reliability of cited articles, so here are a few examples of cites of bnnbreaking.com added in the last few days. Fico Puricelli cited it on G.I. Joe. SiniyaEdita cited it on List of power stations in Sierra Leone. Cassiopeia cited it on UFC Fight Night 244. Chelsdog cited it on Dujuan Richards. (I think there are others.) Peter Gulutzan (talk) 15:55, 16 February 2024 (UTC)

Is this really not the place to discuss the reliability of this source in general?

Anyway, looking at those examples:

[41] looks like a rehash of [42], with the addition of some generic hallucination about the lore of the Transformers universe.
[43] is a close yet incorrect paraphrase of a press release here: [44] (note that commencement of construction is Q2 2024 in the press release but is "the final quarter of 2024" in the BNN article).
[45] is an obvious ripoff of [46]: search for the phrase hungry Argentine up-and-comer to find various syndications of this story, all attributed/credited. All except BNN, that is. I guess their AI model was unable to find a paraphrase for this unusual construction?
Meanwhile, on [47], it appears their paraphrasing extends to the content of literal quotes from an interview, resulting in fabricated quotes attributed to the subject. You can read the original here: [48].

Whoever runs BNN better have good lawyers, as their tech appears to be on an out-of-control copyright infringement rampage. I note they are already facing legal action for defamation as a result of one of their publications.

In the meantime, this source has no place on Wikipedia. I suggest every citation of BNN needs to be removed immediately, possibly replaced with {{Citation needed|date=February 2024|reason=Unreliable source removed}}.

I expect editors using this source will have been acting in good faith, as the BNN articles do look very convincing. Barnards.tar.gz (talk) 23:51, 16 February 2024 (UTC)

I have opened a request to add both bnnbreaking.com and bnn.network to the blacklist: MediaWiki talk:Spam-blacklist#bnnbreaking.com. I’m not sure if this can be actioned while we still have about 100 links to bnnbreaking.com and about 200 to bnn.network. I removed a large number yesterday but more are being added. Help pruning these would be appreciated. Barnards.tar.gz (talk) 23:14, 18 February 2024 (UTC)

In the ongoing discussion regarding bnnbreaking.com, it is imperative to address concerns that some contributors may be engaging in original research or allowing bias to influence their evaluation of the site's reliability and its contributions to Wikipedia. Wikipedia's policies on No Original Research (NOR) and Neutral Point of View (NPOV) are foundational principles that ensure content is verifiable and unbiased, reflecting a fair representation of all significant viewpoints.

While criticisms of bnnbreaking.com have been presented, including accusations of AI-generated content and questionable reliability, it's crucial that these assessments are grounded in verifiable evidence and not the product of conjecture or extrapolation beyond what sources explicitly state. The reliance on indirect indicators or affiliations with other entities to judge bnnbreaking.com's credibility risks veering into the realm of original research, which Wikipedia strictly prohibits.

Moreover, the discourse around potentially blacklisting bnnbreaking.com or removing its citations from articles must be carefully navigated to ensure it does not become an avenue for targeted actions that could be construed as attacking the site. Such actions need to be scrutinized for compliance with Wikipedia's policy against bias and ensuring that discussions and decisions are based on a balanced evaluation of the source's use and contributions to the encyclopedia.

The presence of references from bnnbreaking.com in Wikipedia, especially if cited by reputable news organizations, underscores the need for a nuanced discussion about its role and reliability as a source. It's vital to differentiate between content that may not meet Wikipedia's reliability standards and content that does, ensuring that decisions to remove or retain references are made transparently and with broad community input.

I urge a return to the core principles of verifiability, neutrality, and the avoidance of original research. Let's ensure that our focus remains on enhancing the encyclopedia's quality and reliability through evidence-based evaluations and consensus-driven decision-making. Such an approach will safeguard against undue bias and promote a more inclusive and balanced representation of sources and perspectives. 49.130.118.20 (talk) 14:53, 19 February 2024 (UTC)

P.S. I have also reinstated the 'Criticism section' template as of February 2024 due to unresolved issues with article's balance and neutrality.

Any editor rushing to black list the site and also keep the page in its current form, is clearly NOT abiding by the Wikipedia's NPOV foundational principles. 49.130.118.20 (talk) 14:57, 19 February 2024 (UTC)

On NOR: This policy does not apply to talk pages and other pages which evaluate article content and sources, such as deletion discussions or policy noticeboards.

It is normal and expected that editors will research the reliability of sources.

Please declare whether you have a conflict of interest regarding BNN. Barnards.tar.gz (talk) 15:14, 19 February 2024 (UTC)

Thanks, @Barnards.tar.gz, for highlighting the scope of the NOR policy on talk pages. While it's clear that evaluating source reliability is within bounds, it's also crucial that such evaluations are rooted in verifiable facts rather than conjecture. Discussion should not serve as a pretext for removing content or sources based on unverified assumptions.

Regarding the conflict of interest query: No, I have no conflict with BNN. My interventions are aimed at ensuring a balanced and fair discussion in line with Wikipedia's commitment to neutrality and verifiability. How about you? Any conflicts we should know about to maintain transparency? 49.130.118.20 (talk) 15:35, 19 February 2024 (UTC)

Also, I was very surprised to see the founder's negative history highlighted prominently on BNN Breaking's page which has nothing to do with the subject matter?

@DanielMichaelPerry can you confirm you have no conflicts either? 49.130.118.20 (talk) 15:38, 19 February 2024 (UTC)

The founder’s negative history is not “highlighted prominently”, it is briefly mentioned at the end of the article.

I can confirm I have no conflict of interest. However, the fact that your IP address (along with those of other editors who have tried to remove negative coverage) is registered in Hong Kong - where BNN Breaking is based - strongly suggests that you are not being honest your potential conflict of interest. DanielMichaelPerry (talk) 15:51, 19 February 2024 (UTC)

No, I have no conflict with BNN. Please forgive my fussiness, but I want to make sure you understood the question. I didn't ask if you had a conflict with BNN. I asked if you had a conflict of interest: Conflict of interest (COI) editing involves contributing to Wikipedia about yourself, family, friends, clients, employers, or your financial and other relationships.

I have no conflict of interest. My interest is in protecting the integrity of Wikipedia by rooting out dubious sources. I hope you'll agree that's why we are here having this discussion on the Reliable Sources Noticeboard.

it's also crucial that such evaluations are rooted in verifiable facts I agree. If you read further up the thread, you'll see it's a verifiable fact that BNN has engaged in copyright infringement, so as well as reliability concerns, this site should not be used on Wikipedia per WP:COPYVIO. Barnards.tar.gz (talk) 15:53, 19 February 2024 (UTC)

In addressing the points raised: @Barnards.tar.gz and @DanielMichaelPerry:

Regarding the Founder's History: The selective emphasis and lack of relevance on the founder's negative history, without a balanced view that includes positive achievements, deviates from Wikipedia's guidelines on biographies of living persons (BLP). Wikipedia is not a repository for only negative information. If the founder's history is relevant, then a balanced approach that includes both positive and negative aspects should be applied, similar to how other public figures' pages are structured. This ensures compliance with the NPOV policy.

Conflict of Interest Accusations: My geographical location or IP address does not inherently indicate a conflict of interest. Wikipedia's guidelines on COI are clear about the nature of conflicts, focusing on direct relationships rather than geographical proximity. Accusations based on location alone are insufficient and divert attention from the substantive issues of content neutrality and source reliability.

Copyright Infringement Claims: The assertion that BNN has engaged in copyright infringement relies on verifiable evidence, not just claims or allegations. Wikipedia's standards for inclusion require that information, especially of a negative nature, be well-documented and proven in reliable sources, preferably through legal adjudication or similar authoritative confirmation. Without clear, court-verified instances of copyright infringement, we risk propagating unverified claims, contrary to Wikipedia's verification policy.

In pursuit of maintaining the integrity of Wikipedia, it's essential that our discussions and actions are guided by the platform's core policies: Verifiability, NPOV, and No Original Research. I hope you all can focus on constructive dialogue that seeks to enhance the article's accuracy and neutrality, respecting Wikipedia's standards and the broader community's trust in the encyclopedia. 49.130.9.100 (talk) 16:21, 19 February 2024 (UTC)

You didn't answer the question: Do you have a WP:Conflict of Interest regarding BNN Breaking? Gråbergs Gråa Sång (talk) 16:25, 19 February 2024 (UTC)

@Gråbergs Gråa Sång, you continue to engage in disruptive editing and have not disclosed if you are getting paid for these disruptive practices. https://en.wikipedia.org/enwiki/w/index.php?title=BNN_Breaking&action=history

Knowing how "active" you are on Wikipedia with these disruptive edits, you clearly know I have answered the question in the talk page and have no conflicts. https://en.wikipedia.org/wiki/Talk:BNN_Breaking#Shame_the_site_is_under_attack. 49.130.9.100 (talk) 16:33, 19 February 2024 (UTC)

Not paid, WP is my hobby. And I don't see any comment on your COI on that talkpage, though it's getting a bit long. Perhaps you can quote that statement here? Gråbergs Gråa Sång (talk) 16:42, 19 February 2024 (UTC)

The assertion that BNN has engaged in copyright infringement relies on verifiable evidence, not just claims or allegations. Bingo. Barnards.tar.gz (talk) 16:26, 19 February 2024 (UTC)

Copyright Infringement Claims: The assertion that BNN has engaged in copyright infringement relies on verifiable evidence, not just claims or allegations. Wikipedia's standards for inclusion require that information, especially of a negative nature, be well-documented and proven in reliable sources, preferably through legal adjudication or similar authoritative confirmation. Without clear, court-verified instances of copyright infringement, we risk propagating unverified claims, contrary to Wikipedia's verification policy. 49.130.9.100 (talk) 16:34, 19 February 2024 (UTC)

No. We don't need a court judgement to note that a site contains copyright violations and we shouldn't link to it. MrOllie (talk) 16:36, 19 February 2024 (UTC)

Sure, so if any random user with a userId starts complaining about a site that reaches millions of users, and claims it has a copyright violation, we should these userIDs as reliable sources?

You are not a lawyer, nor are you a journalist. I suggest you stick to your adherence to Wikipedia policies. Which apparently you don't understand. 49.130.9.100 (talk) 16:49, 19 February 2024 (UTC)

Discussions on a talk page are not held to the same sourcing standards as claims within articles. By your standard policies such as WP:COPYLINKS would not be enforceable - but I assure you they are. MrOllie (talk) 16:58, 19 February 2024 (UTC)

Instead of engaging in personal attacks (which will eventually get you blocked), it would be more helpful if you could explain why the cited examples of copyright infringement on your site are not actually copyright infringement. That would be the easiest way to make the problem go away. Barnards.tar.gz (talk) 17:07, 19 February 2024 (UTC)

Mr IP - you should try to sound less like AI generated text yourself when you are defending a source who has been accused of generating their content with AI. MrOllie (talk) 16:34, 19 February 2024 (UTC)

My apologies my thoroughness has bothered you. I suggest you try to incorporate some happiness into your life before replying. 49.130.9.100 (talk) 16:47, 19 February 2024 (UTC)

IP user is now blocked for edit warring--obviously. Drmies (talk) 17:09, 19 February 2024 (UTC)

After seeing the harassment here, and the other edits from similar IPs, I placed a rangeblock. Geolocation is Hong Kong, for what it's worth. Drmies (talk) 17:14, 19 February 2024 (UTC)

See also Wikipedia:Reliable sources/Noticeboard/Archive 421#BNN Breaking. This is the only prior discussion I've found. WhatamIdoing (talk) 18:26, 19 February 2024 (UTC)

And if anyone missed it, now blacklisted. Initially I thought that suggestion was a little hasty, but BNN changed my mind. Gråbergs Gråa Sång (talk) 18:32, 19 February 2024 (UTC)

Rudolf Steiner and Anthroposophy

This is about [49] and [50].

The sources are:

Taverne, Dick (2006). The March of Unreason: Science, Democracy, and the New Fundamentalism. OUP Oxford. p. 28. ISBN 978-0-19-157861-8. Retrieved 3 February 2022. Rudolf Steiner joined the Nazi party in its early days—supports the claim that Steiner was a member of the NSDAP.
Garner, Richard (24 January 2007). "The Big Question: Who was Rudolf Steiner and what were his revolutionary teaching ideas?". The Independent. Retrieved 26 October 2023.—supports the claim that Hitler declared "war against Steiner".

Was discussed at Talk:Rudolf Steiner#Drop the claim. tgeorgescu (talk) 06:20, 17 February 2024 (UTC)

On Taverne, it is OUP; however, in checking the book, Taverne twice claims Steiner joined the Nazi Party in the early days but neither claim is supported by a reference (Steiner died in 1925 so it would have to be very early days). I also checked the nearest footnote in case it covered more (Bramwell, Anna (1989-01-01). Ecology in the 20th Century: A History. Yale University Press. ISBN 978-0-300-04521-5.) and though it covers Rudolf Steiner and his movement extensively it does not mention him joining the Nazi party (as far as I can see, I don't have full access) though it does mention that his ideas influenced some prominent members of the Nazi party (and not others). Taverne also freely admits he is not a trained historian. For this particular claim I would say not reliable.

On Garner, this is a newspaper opinion piece including history. I would say there should be far better sources. Erp (talk) 15:31, 17 February 2024 (UTC)

As far as I could investigate the matter, "war against Steiner" was a statement by a German Catholic nationalist, in a Catholic nationalist newspaper. I saw no evidence that Hitler ever said something like that. It is not odd for Catholic true believers to lambaste heretical movements. On the other hand, Hitler was not a Catholic true believer, and he did not really care about heresy. tgeorgescu (talk) 20:01, 17 February 2024 (UTC)

I think the sourcing for the claim that Steiner joined the Nazi party is weak enough that the claim should be omitted. Hemiauchenia (talk) 00:05, 20 February 2024 (UTC)

About NewsMeter, Newsmeter, newsmeter.in and so on used as references

Hello all,

I can see that NewsMeter, Newsmeter, newsmeter.in and so on have never been created.

Those or variants are currently used on

If appropriate, please do ask for or perform WP:REVDEL on any or all of this without asking me.

I seek your opinions about this.

Shirt58 (talk) 🦘 10:33, 20 February 2024 (UTC)

Acceptable use of SPS?

A self-published academic study on the subject of transport makes an important point, published (on Google Docs) in a form acceptable to policy, by a multiply-published expert, see [51] and [52]. I feel that we should include it in the article Low Traffic Neighbourhood, as follows:

In 2022 local elections were held across London. Many candidates had tweeted about LTNs, Labour candidates generally positively, Conservatives generally against; these tweets seemed to make very little difference to the number of votes cast for these candidates. If anything, tweeting positively about LTNs may have increased the number of votes for Labour councillors.Sound and fury? The impact of councillors’ LTN positions on voting behaviour in Greater London. July 2023. Jamie Furlong, Athena Brook, Charlie Hicks, Professor Rachel Aldred. accessed 28 Jan 2024.

Is this source acceptable? Richard Keatinge (talk) 14:04, 12 February 2024 (UTC)

Google docs are not stable. Bon courage (talk) 14:18, 12 February 2024 (UTC)

This is the point about our definition of publication and our criteria for the use of SPS. This Google Doc seems entirely acceptable according to policy, although, as with most other SPSs, the author could change it at any time. Richard Keatinge (talk) 15:26, 12 February 2024 (UTC)

One can usually take a snapshot. But why would we use a self-published Google Doc for a topic where there are oodles of peer-reviewed academic sources avialable? If the paper is truly making "an important point" surely RS will have made that point? Bon courage (talk) 15:44, 12 February 2024 (UTC)

This particular point is significant in view of the electoral significance and social media uproar about LTNs, and this is the only paper I've found that makes it. The author, Professor Aldred, tells me: "We're hoping to publish this in a journal but it's currently not yet been submitted so given academic publication timeframes, it might not be for a while. Sorry." Richard Keatinge (talk) 15:53, 12 February 2024 (UTC)

I would say wait until it IS published… in the meantime, keep looking for another source. Blueboar (talk) 16:15, 12 February 2024 (UTC)

OK...could we have your opinion on whether it's an acceptable source according to policy? I think it unequivocally is. If so I'd hope then to develop a consensus for or (more likely given comments so far) against its use, on the appropriate talk page. Richard Keatinge (talk) 16:44, 12 February 2024 (UTC)

Everything's reliable for what it says (and this wouldn't be usable for assertions of fact). But the problem more is NPOV. Why would Wikipedia be paying attention to this source when no RS has? Wikipedia is meant to be a tertiary source, not an ersatz secondary one picking winners from among unpublished primary sources. Bon courage (talk) 17:29, 12 February 2024 (UTC)

Can't he put it in a preprint directory or something? We do, for example, have a {{cite arxiv}} template. jp×g 🗯️ 17:56, 20 February 2024 (UTC)

Not that this is the Most Impeccable Triumph Of Sourcing Excellence, of course, but you could do worse than arxiv to source research. jp×g 🗯️ 18:13, 20 February 2024 (UTC)

Wikipedia policy is that Self-published expert sources may be considered reliable when produced by an established subject-matter expert, whose work in the relevant field has previously been published by reliable, independent publications (see WP:EXPERTSPS). As co-authors Jamie Furlong and Rachel Aldred are both established experts in the subject of Low Traffic Neighborhoods who have been published in reliable academic sources, on those grounds, this self-published source is a reliable source.

Whether or not the proposed content should be included is a separate question of content, to be resolved at the talk page for the article. P-Makoto (she/her) (talk) 19:34, 12 February 2024 (UTC)

What evidence is there this was published by her? Slatersteven (talk) 19:42, 12 February 2024 (UTC)

This has been published via the wearepossible.org website. See their article[53] and it's link to the full report in pdf format[54] I don't see any reason to use the Google docs version. -- LCU ActivelyDisinterested «@» °∆t° 22:34, 12 February 2024 (UTC)
Then no the google docs version is not an RS, as we do not know who published it, there are more "reliable" versions out there. Slatersteven (talk) 11:20, 13 February 2024 (UTC)

Also here on Westminster Research via the university of Westminster. It looks like that just publishes any research by the University of Westminster's academic community, and its about page clearly states that The administrator only vets items for the eligibility of authors/depositors, and valid layout & format. The validity and authenticity of the content of submissions is not checked. so it still counts as a WP:SPS, but given that it clearly passes WP:EXPERTSPS it is at least notionally usable for a brief relatively-uncontroversial sentence like this. --Aquillion (talk) 14:23, 16 February 2024 (UTC)

Thanks to all contributors. At this stage, would anyone disagree with Aquillion's "notionally usable" verdict? If not, it's time to wrap this discussion up. Richard Keatinge (talk) 17:10, 18 February 2024 (UTC)

[1] Hassan, Chris (August 14, 2023). "X-Men Monday #215 – Gerry Duggan Talks 'X-Men' at FAN EXPO Boston 2023". AIPT Comics. Retrieved February 18, 2024.

[2] Johnston, Rich (February 18, 2024). "Marvel Cancels All X-Men Comics in the Daily LITG, 18th February 2024". Bleeding Cool. Retrieved February 19, 2024.

[3] Johnston, Rich (February 7, 2024). "Invasion, Insurrection & Kissing in Today's X-Men Comics (XSpoilers)". Bleeding Cool. Retrieved February 19, 2024.

[1]

[2]

[3]