Wikipedia:Reliable sources/Noticeboard/Archive 395

In one of the sources provided by Foorgood (F. Lionel Young, III, World Christianity and the Unfinished Task: A Very Short Introduction, Wipf and Stock, 2020), which is itself a book dedicated to a particular Protestant missionary project and view, you can read the following lines: ...Barrett's research has continued under the auspices of an organization established in 2001 named the Center for the Study of Global Christianity, now situated on the campus of Gordon-Conwell Theological Seminary. The center's co-director, Todd Johnson, began working with Barrett in 1989, and collaborates with his colleague on several projects, including the 2001 edition of the WEC. Building on Barrett's groundbreaking work, the center launched the World Christian Database and the World Religion Database....

As a general example and point of reference, compare ARDA projections about Australia to the Australian 2021 Census (ARDA overestimates Christianity by 14%); ARDA projections about Canada to the Canadian 2021 Census (ARDA overestimates Christianity by 10%). They are completely wrong, for every single country.

A further critical remark is that ARDA data are speculative projections, not actual surveys, and therefore violate

• Liedhegener A.; Odermatt A. (2013). "Religious Affiliation in Europe - an Empirical Approach. The "Swiss Metadatabase of Religious Affiliation in Europe (SMRE)". Zentrum für Religion, Wirtschaft und Politik (ZRWP),
  doi:10.13140/RG.2.2.33430.55364
  .

• Hsu, Becky; Reynolds, Amy; Hackett, Conrad; Gibbon, James (2008). "Estimating the Religious Composition of All Nations: An Empirical Assessment of the World Christian Database" (PDF). Journal for the Scientific Study of Religion. 47 (4): 691–692.
  doi:10.1111/j.1468-5906.2008.00435.x
  .

• Woodberry, Robert D. (2010). "World Religion Database: Impressive - but Improvable". International Bulletin of Missionary Research. 34 (1): 21–22.
  ISSN 0272-6122
  .
  • Quote: ... 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 ....
  • However, context matters: 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.

• Marsh, Christopher; Zhong, Zhifeng (2010). "Chinese Views on Church and State". Journal of Church and State. 52 (1): 34–49.
  JSTOR 23922246
  .
  • Quote: overestimate of Christianity in China, which adds a lot to the total number and percentage of world Christians: ... At the extreme high end, the World Religion Database puts the percentage of Christians in China at 7.76 percent, or a just above 100 million, but this number is most certainly an overestimation ....

Another edition of WRD/WCD data has been the Atlas of Global Christianity (produced by the same Gordon-Conwell team). I have found negative critical assessments even for this edition, this time coming from an "insider" (Christian missionary) source, even though through an academic publisher, written by Anne-Marie Kool of the Evangelical Theological Seminary of Osijek, Croatia:

• Kool, Anne-Marie (2016). "Revisiting Mission in, to and from Europe through Contemporary Image Formation" (PDF). In Charles E. Van Engen (ed.). The State of Missiology Today: Global Innovations in Christian Witness. Downers Grove: IVP Academic. pp. 231–49.
  • p. 1: ... [the resource] seeks to give “as nuanced a picture as possible” of the history of Christianity over the last 100 years showing an “unmistakable” general pattern, that Christianity experiences a “severe recession” on the European continent that once was its primary base, while it has undergone “unprecedented growth and expansion” in the other parts of the world. ....
  • p. 2: ... widespread caution is raised with regard to the accuracy of the figures and not to engage in statistical analysis with the data, “without robustness checking… they contain random error and probably some systematic error” ....
  • p. 9, containing a self-criticism from Kool for having herself made uncritical use of the data: ... The World Christian Database and the World Religion Database serve as sources for the data of the Atlas. With regard to the methodology used, Woodberry is right in emphasizing that “more transparency is needed”. 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. 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. ....
  • p. 12, about systematic overestimation of Christianity in Europe, with allusions that there might be financial reason behind such overestimations: ... The statistical image of Europe that is now communicated only re-enforces the image of Europe as a Christian continent, by not giving insight in the internal diversification and erosion. So why is only this broad definition used? Is it for fear of losing power? Or for maintaining the image of the numerically strong “World C” that provides the human and financial resources to “finish the task”? Are matters of Christian finance playing a role? Out of a sense of empire building? Or of a sense of hidden resistance to accept that Europe also is now also a mission field? Is it out of fear of becoming a minority? Fear for ending up statistically weaker than the Muslims? Or an attempt to cling to the influence of the “Western” over the “non-Western” world, based on an image of Europe as still a massive Christian continent? ....
  • p. 13; it is a missionary tool designed for a specific strategy of aggression towards what in American missionary Christianity has been conceptualised as the "10/40 window": ... Eric Friede’s sharp analysis points us to the fact that the Atlas is ultimately written from the perspective of the so-called Great Commission Christians, Christians who engage in and support Christian missions, as many essays address the issue of “how to grow Christianity” in a particular region. The mission strategy invoked is then one of identifying within Global Christianity the resources needed for the task, the human resources, the GCC Christians, as well as the Christian finances that could make this enterprise work. An assessment of major tools needed for finishing this task is offered in subsequent sections, like Bible translation is followed by a section on Evangelization, with a division of the world in A, B and C, according to the level “being evangelized”. Statistics are used to motivate missionaries and national workers to mission action with Christian mission being reduced to a manageable enterprise with a dominant quantitative approach and a well-defined pragmatic orientation, “as a typical school of thought coming from modern United States”. .... Kool makes largely reference to: Eric Friede, "Book Review. Atlas of Global Christianity: 1910-2010, by Johnson, T.M. & Ross, K.R. (Eds.), 2010", Edinburgh, Edinburgh University Press, Theological Librarianship 3(1), 2010.

Among other sources, some of which we have already analysed, Kool makes reference to:

• Brierley, Peter. (2010). "World Religion Database: Detail Beyond Belief!". International Bulletin of Missionary Research. 34 (1): 18–20.
  • It is a critical assessment, once again coming from a missionary journal, that raises doubts as to the reliability of the WRD on the basis of the mismatch of the latter's data (purportedly based on censuses) with actual data from censuses, in particular those of the UK. Judd Birdsall and Lori Beaman, in Faith in Numbers: Can we Trust Quantitative Data on Religious Affiliation and Religious Freedom?, Transatlantic Policy Network on Religion & Diplomacy, 22 June 2020, at p. 3 say that the WRD, despite being widely cited and impressive, "comes with limitations. In his review of the Database, the statistician Peter Brierley pointed out that for the United Kingdom the Database used denominational reports, such as Church of England baptismal records, rather than the UK census figures to calculate affiliation. A tally of denominational reporting showed that 82% of Britons were Christian, whereas only 72% of them claimed to be Christian in the UK census". (n.b. Brierley makes reference to the UK 2001 census data, showing that already in 2001 the WRD overestimated Christianity in the UK by at least 10%).
  • (Added by Ramos1990). However context matters. The same source states: 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.

Discussion (ARDA and WRD)

Part 1

The World Religion Database provides its estimates based on census and surveys: https://www.worldreligiondatabase.org/, just like Pew Research does. The sources I cited above from Oxford and Princeton call it very reliable and accurate even though it is not exact as Censuss but estimates like Pew are used all over Wikipedia.Foorgood (talk) 00:39, 29 October 2022 (UTC)

That being said, I want to confirm whether or not AEOs position is that ARDA/WRD estimates shouldn't override national censuss(which I agree with) meaning they could still be used for other estimates, OR if AEOs position is that ARDA/WRD should not be used at all(which is absurd given their reliable reputation with Oxford and Cambridge)?Foorgood (talk) 01:04, 29 October 2022 (UTC)

The WRD is not a census (survey of the entire population of a state by that state's statistical office), and, as demonstrated by Erp herebelow, their methods for collecting and elaborating data are unclear (and, n.b.,

WP:DEPRECATED. Æo (talk

) 10:47, 29 October 2022 (UTC)

Nope you cannot radically decide to block PEW and ARDA, both globally recognized as top reliable sources, from Wikipedia just because you now think you know better than them. But what we can do is give preeminence to Censuss while allowing estimates to be provided lower in the article with the disclaimer that they are not official surveys etc. Foorgood (talk) 14:03, 29 October 2022 (UTC)

"Both globally recognized as top reliable sources". Please note that such alleged "global recognition" is basically the result of their own campaigns of promotion, and support by their allied journalistic media. Take the F. Lionel Young source cited above: it indeed praises the WRD within a chapter dedicated to statistical sources which are part of a precise Protestant Christian missionary project. These are, very simply, unreliable biased sources. Obviously, I cannot classify them as deprecated myself; this would require community consensus. Let's see how the present discussion will develop before proceeding with further steps. Æo (talk) 14:45, 29 October 2022 (UTC)

You actually think that PEW or ARDA, globally recognized as reliable, would meet all the requirements listed here?!: https://en.wikipedia.org/wiki/Wikipedia:Deprecated_sources.. Again are you so extreme in your stance that you can't come to a compromise like you've done already by simply having Censuss take top priority on nations religions pages? Foorgood (talk) 15:26, 29 October 2022 (UTC)

Regarding census data I have not come to any compromise; census data are simply the best, most accurate available. And yes, I think ARDA/WRD/WCD (alias Gordon-Conwell) and Pew-Templeton Global Religious Futures (projections based on outdated 2001-2010 surveys/collections of data) meet the requirement for deprecation. Note that deprecation does not mean banning a source (blacklisting), it's just a warning that will appear whenever contributors will insert links to such sources. Æo (talk) 17:10, 29 October 2022 (UTC)

Ok so you are saying that you don't feel the PEW or ARDA data needs to banned entirely from articles, just given disclaimer that it's not an official survey like a census? Foorgood (talk) 17:34, 29 October 2022 (UTC)

ARDA itself is a database archive and states that when citing ARDA the original source must be included in the cite so the key question in most cases is the reliability of WRD. I have access to the World Religion Database so decided to check one recent cite in the Demographics of atheism article where the claim was "In 2020, the World Religion Database estimated that the countries with the highest percentage of atheists were North Korea and Sweden". First how on earth does anyone know what the percentage of the population are atheists in North Korea? Tunneling down through the WRD yielded the source for its info on religion in that country as "North Korea, Future of the Global Muslim Population (FGMP), 2020" and a note at the bottom of a fairly blank page was "Pew Forum Projection". Unfortunately the Pew FGMP (a) doesn't mention atheists and (b) cites the WRD as its source for the Muslim population of North Korea. I do note a WRD discussion of its methodology is at https://worldreligiondatabase.org/wrd/doc/WRD_Methodology.pdf including the paragraph:

"Religious demography must attempt to be comprehensive. In certain countries where no hard statistical data or reliable surveys are available, researchers have to rely on the informed estimates of experts in the area and subject. Researchers make no detailed attempt at a critique of each nation’s censuses and polls or each church’s statistical operations. After examining what is available, researchers then select the best data available until such time as better data come into existence. In addition, there are a number of areas of religious life where it is impossible to obtain accurate statistics, usually because of state opposition to particular tradition(s). Thus it will probably never be possible to get exact numbers of atheists in Indonesia or Baha’i in Iran. Where such information is necessary, reasonable and somewhat conservative estimates are made."

My suspicion is the estimate of the number of atheists in North Korea is a guess with very large error bars. The number of atheists in Sweden is likely to be more accurate though the latest survey they used is 2008. One should check what definition of atheist is being used by WRD. Erp (talk) 01:31, 29 October 2022 (UTC)

Excuse me the 2017 Win/Gallup poll also has Sweden as the 2nd most atheist country and here https://en.m.wikipedia.org/wiki/Religion_in_North_Korea we can see 2 different estimates are used to give North Korea's high atheist percentage. Estimates have their own methodology and they are considered Reliable by Oxford and Cambridge so don't try to reinvent the wheel and say that we know better than these statisticians because if so you're going to have to remove every single estimate on Wikipedia for every topic- and there are thousands. Our job on Wikipedia is to include estimates that are reliable while obviously giving precedence to government surveys *When available*.Foorgood (talk) 01:52, 29 October 2022 (UTC)

Actually the estimates for North Korea in the Religion in North Korea are for 'no religion' which is not the same as being atheistic. Also the 2017 Win/Gallop report (https://web.archive.org/web/20171114113506/http://www.wingia.com/web/files/news/370/file/370.pdf) has China as being the 'least religious', not atheistic, of the countries polled with Sweden second. However in the same report when it comes to percentage stating they are atheists Sweden drops below China, Hong Kong, Japan, Czech Republic, Slovenia, South Korea, France, and Belgium. North Korea for obvious reasons was not among the countries polled. This does not help in showing that WRD is a reliable source. Note that does not mean I agree with @Æo that censuses are the best sources; censuses can have biases or be incomplete and good surveys/polls can be just as reliable or better if not as precise. I would be happier with WRD if it were specific on how it got its figures for each country (among other things it would avoid articles citing X and then citing WRD which in turn was using X). Erp (talk) 17:38, 29 October 2022 (UTC)

If you click here https://www.thearda.com/world-religion/np-sort?var=ADH_704#S_1 and read the top it says "Variables: Total number of Atheists by country and percent of population that are Atheists: Persons who deny the existence of God, gods, or the supernatural. (World Religion Database, 2020) (Atheists)1" Again, you guys are acting like you know better than these world renowned sources. Foorgood (talk) 17:43, 29 October 2022 (UTC)

Foorgood, I have thoroughly demonstrated that such "world renowned sources" are not produced by actual statisticians but by Protestant missionaries and Erp has demonstrated that their methods for collecting data are dubious. The line you have cited does not mean anything as to statistical survey methodology, it is just a conceptual category they have used to represent their data. Please read

WP:NOTHERE. Æo (talk

) 23:33, 29 October 2022 (UTC)

I suspect the precision of exactly 4,016,422 atheists in North Korea in 2020, too precise for what is suppose to be an estimate. Have the authors not heard of significant figures or more likely it is an issue with the database design not being able to round? Also the definition at the top of the ARDA WRD chart is not quite the same as the one in the World Religion Database (the numbers do match). The latter definition is "Number of Atheists in this country's population. Atheists are persons professing atheism, skepticism, impiety, disbelief or irreligion, or Marxist-Leninist Communism regarded as a political faith, or other quasi-religions, and who abstain from religious activities and have severed all religious affiliation; and others opposed, hostile or militantly opposed to all religion (anti-religious); dialectical materialists, militant non-believers, anti-religious humanists, skeptics." There is a separate category for agnostics.

As for world renowned? Something can be well known yet still not be deemed reliable. I did a search of the Wiley Online Library for "World Religion Database", 27 hits though 16 of them were to a single book by the people who created the database. Wiley also includes the Journal for the Scientific Study of Religion which had two of the cites (one of which was a critique of the World Christian Database). One would think people contributing to a journal on the scientific study of religion might be using this database extensively? I also did a search on "Stanford Encyclopedia of Philosophy" in Wiley, that had 1,248 hits.

I will note that Brian Grim's background does include a PhD in sociology from Penn State which should ensure some statistical training. Erp (talk) 00:35, 30 October 2022 (UTC)

Oxford and Cambridge and 3 other top publishers call them reliable 12, 3, 4, 5 Why? Because they are all statisticians from Universities around the country: ARDA Advisory Board: Renata Curty (UC Santa Barbara), Joel Herndon (Duke University), Nathaniel Porter (Virginia Tech), Ruth Tillman (Pennsylvania State University), Andrew Tyner (Center for Open Science)

ARDA Affiliates: US Religion Census, Baylor Univeristy, World Religion Database at Boston University, which is part of Brill publishing: https://www.worldreligiondatabase.org/

No, you do not know better than these experts. But AEO I'm asking you, your position is that ARDA not be banned from articles you just want it with the deprecated tag?Foorgood (talk) 01:16, 30 October 2022 (UTC)

And ERP you will have to show us a screenshot if you think we will simply take your word about what WRD classifies as atheism because there is so far absolutely 0 proof of what you just stated. Foorgood (talk) 01:44, 30 October 2022 (UTC)

@Foorgood I'm not sure I'm permitted to put a screen shot in this discussion (wikipedia images are suppose to be stuff we can use in articles) or even if it would be sufficient proof for you given you apparently have no access to the database and therefore don't know what it looks like (I could after all photoshop it). Would it be better to have a third party who has access to WRD to vouch for the accuracy? A party you choose. I'm not sure whether @NebY or @Æo have access. Erp (talk) 04:24, 31 October 2022 (UTC)

Nope because as I wrote below: Here is The Harvard Library calling World Religion Database "a good source of statistics" https://guides.library.harvard.edu/religion and here's The Stanford Library https://guides.library.stanford.edu/religion saying of Arda "Data included in the ARDA are submitted by the foremost religion scholars and research centers in the world. The ARDA allows you to interactively explore the highest quality data on American and international religion using online features for generating national profiles, GIS maps, church membership overviews, denominational heritage trees, tables, charts, and other reports." University of Oxford Library also recommends both of them https://www.bodleian.ox.ac.uk/collections-and-resources/data/finding-data/themes/religion. All the top Universities call it a reliable source period you do not have ANY reputable source that calls it unreliable it is completely your personal opinion from your own original research. Foorgood (talk) 14:03, 31 October 2022 (UTC)

@Erp: I don't have access to the database, but I personally trust your word. In any case, I think it would not be a problem if you uploaded a little screenshot of the section of the page which demonstrates unclear and circular reporting; I think it would not be a copyright violation. Regarding Foorgood, I think he is gaming the discussion system by bringing the interlocutor to exhaustion, ignoring the evidence we have put forward and stubbornly copy-pasting his links which do not demonstrate anything except that ARDA/WRD is listed among other sources on some university/library websites. Æo (talk) 17:05, 31 October 2022 (UTC)

I would also remind @

WP:AGF. We have no reason to suspect Erp of fabricating a quotation from WRD and I very much hope that FoorGood doesn't imagine that such suspicion would be justified by or would justify misleading statements by Foorgood themself. NebY (talk

) 17:41, 31 October 2022 (UTC)

Many thanks for pointing that out about Good Faith. I would like someone else to check given that I am using an older browser so perhaps some things are hidden from me such as lack of sources (not the definition of atheists, that is definitely there). Or I should check on a different computer. I note @Foorgood has been contacting various people and one of them might have access.

I'm actually not so sure it was circular reporting since it isn't clear whether the surveys listed by WRD were actually listed sources or listed links for related information. The idea behind the WRD makes a certain amount of sense; however, the methodology is lacking in a few ways. What are the sources for each country and a short description on how they are used, who is responsible for the calculations in each country (or are the listed editors, Todd M. Johnson, Brian J. Grim, Gina A. Zurlo, Peter Crossing, and David Hannan, responsible for all countries?), are there regular archives so a researcher using it doesn't find the data changing out from underneath them (these archives might exist); why aren't figures rounded to avoid giving a precision that is impossible for estimates? what are the error estimates?

By the way if WRD is well known (whether for good and/or for bad), it probably should have its own Wikipedia article. Erp (talk) 02:01, 1 November 2022 (UTC)

Foorgood, your wrote "Oxford and Cambridge and 3 other top publishers call them reliable" and previously "they are considered Reliable by Oxford and Cambridge". Books publishesd by Oxford University Press and Cambridge University Press don't represent the opinion or judgment of OUP or CUP (or of the universities of Oxford and Cambridge), and this is true of publishers generally; for example, a book published by

Harper Collins does not represent the opinion of Rupert Murdoch. An advisory board of statisticians "from Universities around the country" isn't automatically of high quality (they might be the best in the US or they might be the only ones in the US who'll work with that organisation) and the extent to which advisory boards influence an organisation's work and output varies massively. NebY (talk

) 14:13, 30 October 2022 (UTC)

Here is even The Harvard Library calling World Religion Database "a good source of statistics" https://guides.library.harvard.edu/religion and here's The Stanford Library https://guides.library.stanford.edu/religion saying of Arda "Data included in the ARDA are submitted by the foremost religion scholars and research centers in the world. The ARDA allows you to interactively explore the highest quality data on American and international religion using online features for generating national profiles, GIS maps, church membership overviews, denominational heritage trees, tables, charts, and other reports." University of Oxford Library also recommends both of them https://www.bodleian.ox.ac.uk/collections-and-resources/data/finding-data/themes/religion. All the top Universities call it a reliable source period you do not have ANY reputable source that calls it unreliable it is completely your personal opinion from your own original research. Foorgood (talk) 16:01, 30 October 2022 (UTC)

Foorgood, note that the sites you have linked merely list ARDA among other sources, and the blurb is likely a self-presentation. They are not critical researches; critical assessments are those like the ones from which I have excerpted the quotes reported at the beginning of the discussion (Liedhegener et al. 2013, Hsu et al. 2008). Also please note, and I repeat this for the umpteenth time, that the ARDA acquired all its data from the WRD only by 2021/2022, and before then it hosted completely different data. As already expressed before, the first problem here is the WRD, and the ARDA is the secondary problem as it functions as the dissemination platform of WRD data. Æo (talk) 17:54, 30 October 2022 (UTC)

And as you saw Harvard calls World Religion Database a good source so do yourself a favor and stop humiliating yourself trying to make it seem deprecated and pretending you know better than Harvard Stanford and Oxford. Even the Yale and Princeton Library websites suggest World Religion Database. Foorgood (talk) 17:57, 30 October 2022 (UTC)

That's better, at least to start with; you clearly identify those mentions as being from university libraries. But then you veer into saying "all the top universities", as if the libraries are the universities, as if those samples do call it a reliable source, and as if your sample proves that all "top" universities or even their libraries call it a reliable source. And then you tell me that "you do not have ANY reputable source that calls it unreliable it is completely your personal opinion from your own original research". You do not know what my opinion is; my comment above on your statements was my first. NebY (talk) 17:57, 30 October 2022 (UTC)

That first sentence from Stanford Library, "Data included in the ARDA are submitted by the foremost religion scholars and research centers in the world", is also in our article

WP:RS, or both might be taken from a self-description of ARDA, whose website currently has "now including American and international collections submitted by the foremost religion scholars and research centers in the world."[43] (I notice that's not such a strong statement, not making a claim about all the data.) It does not appear to be Stanford Library's independent appraisal of ARDA. NebY (talk

) 17:31, 31 October 2022 (UTC)

Correct, as I wrote above the blurb is a self-presentation copied and pasted here and there, including on Wikipedia. Also note that the self-proclaimed "foremost religion scholars and research centers in the world" are fundamentally the same people of the WRD alias

Gordon-Conwell and of the John Templeton Foundation (another organisation about which we could report plenty of criticism). Æo (talk

) 17:38, 31 October 2022 (UTC)

Sorry, I missed that you'd found it on wikipedia too! NebY (talk) 17:46, 31 October 2022 (UTC)

(Personal attack removed) Foorgood (talk) 18:13, 31 October 2022 (UTC)

I think that much of the evidence presented so far by User:Foorgood indicates that ARDA is a reliable source. The set of references that were provided 12, 3, 4, 5 indicate that there is scholarly usage of such a database. Keep in mind that all major surveys have their limitations and none are really the final word - especially on atheism. Estimates of atheism are particularly problematic e.g. estimates from China differ between surveys (WIN-Gallup International vs Pew Research Center) and China and well... Asia alone shifts the global estimates of atheism considering that just China by itslef has the greatest number of atheists in the world. From wikipedia's stand point there is no issue using ARDA. It is not a depreciated source. It has limitations and problems like all other surveys. Attribution may solve the issues of putting any results from any particular survey in wikipedia's voice.

From the arguments presented against ARDA, none indicate that it is a depreciable source. ARDA has notable sociological researchers like Roger Finke in its board [44] and peer reviewed articles on it are also available [45]. Also, there are many hits from other scholarly sources on google scholar using the database too [46].Ramos1990 (talk) 00:11, 2 November 2022 (UTC)

In my view it is not ARDA that is in question but rather one of the databases it archives, World Religion Database. ARDA itself says that the source should be cited with ARDA just being the repository.

If I modify the google scholar search for just "World Religion Database" I get 559 results though some of them aren't exactly supportive. For instance

one article reviewing the database states

Second, 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

and

[things which] would radically improve the usefulness and face-validity of the data:

• Documenting how each estimate was calculated. A Webbased format is ideal for revealing this kind of information: most users would not be interested in the details, and costs to print such information would be exorbitant.
• Providing some measure of uncertainty with each estimate (e.g., standard errors or even a qualitative evaluation by the editors). Researchers could then integrate uncertainty into their statistical models or exclude cases with uncertain estimates. As it is, estimates for Afghanistan, Algeria, China, and North Korea appear as precise as estimates from Canada and Germany. (Woodberry, Robert D. (2010). "World Religion Database: Impressive - but Improvable". International Bulletin of Missionary Research. 34 (1): 21–22.
  ISSN 0272-6122
  . Retrieved 2022-11-02.)

I also noted that many of the other cites were in articles authored by the database creators.

I also did a search on jstor which tends to be a bit more selective than google scholar on what is scholarly though some recent stuff (3-5 years) may not yet be available on it. There were 31 results (with at least 3 of those by people directly involved in the database). One is by Hsu et al. 2008 mentioned above which is not favorable to the database. One article had the statement

"Relying on the 2010 estimates of the World Religion Database (WRD), this method is used in instances where no better data than the religious composition of the birth country were available" (Henning, Sabine; Hovy, Bela; Connor, Phillip; Tucker, Catherine; Grieco, Elizabeth M.; Rytina, Nancy F. (2011). "Demographic Data on International Migration Levels, Trends and Characteristics". International Migration Review. 45 (4): 979–1016.

JSTOR 23922246

. Retrieved 2022-11-02.)

WRD seems to be on the edge of acceptability. At most it should only be used if no better source exists (I still can't imagine how they calculated the so very precise numbers for North Korea). Erp (talk) 02:48, 2 November 2022 (UTC)

WRD is not on the edge of acceptability. It is clearly used in the sources you cited above. Brill, which is an academic publisher, publishes the database [47] and it comes from Boston University by the way so it is RS by that measure alone. I see no reason to object to it from wikipedia's policy standpoint. The sources you brought do not show that it is a bad database, because clearly peer reviewed sources do use it. Google scholar also produces peer reviewed articles and books that use it too. I got more than 800 when looking at "world religion database (WRD)". How is this a problem for it? It is used quite a bit. Bad sources do not get used this often to build on research. So clearly it has value for academics.

Now, there are studies that do sloppy work on atheism such as WIN-Gallup which showed global atheism rising way too fast in 2012, and magically declining by half in the subsequent WIN-Gallup surveys within the same decade. This of course is preposterous - that atheists would double and then decline in 10 years. People, on a global scale, do not change radically one way and then change back in a decade. And some researchers have advised caution on WIN-Gallup's data set (e.g. Oxford Handbook of Atheism) since their numbers on atheism in China are way too large - compared to all data sets on religion and atheism available. But none of this makes WIN-Gallup an non usable or depreciated source on wikipedia.

That is because all studies have their weaknesses and they usually contradict each other in the literature (Pew vs WIN-Gallup vs WVS vs census data, etc). Wikipedia just presents what certain data sets have come up with. We as wikieditors do not psychologize or make assessments or judgments on how one database is good or bad methodologically. As long as WRD has an academic standing in some way in the literature, there is no real reason to discount it over any other. Attribution should take care of placing the weight on the database being cited for the numbers.Ramos1990 (talk) 06:10, 2 November 2022 (UTC)

As it has already been written and amply demonstrated, the ARDA is a repository of data which at some point between 2021 and 2022 changed all its datasets with those of the WRD/WCD alias Gordon-Conwell. It is the latter that is in question in this discussion, and judgment on the ARDA follows judgment on WRD/WCD. As it has been widely demonstrated, the WRD/WCD is a Protestant Christian encyclopedia, dataset and missionary tool. Its sources are Protestant Christian missionaries, for the most part, as stated in its own methodology paper (pp. 13-14: ... The WRD taps into knowledge from contacts in every country of the world who inform us on what is happening in non-traditional forms of Christianity, such as churches and insider movements ...; notice that some of these firsthand informers, "insiders", are sometimes completely out of reality: for instance, in 2013 some Protestant churches predicted that 10% of Mongolians would be Christians by 2020, yet between 2010 and 2020 (census data) Christians in the country have declined from 2.2% to 1.3%).

As for the sources you have listed (which are the same links provided by Foorgood), the respectability of the publisher or hosting site does not necessarily imply that the content is qualitative and reliable. Cf.

WP:SCHOLARSHIP

: "POV and peer review in journals": Care should be taken with journals that exist mainly to promote a particular point of view. A claim of peer review is not an indication that the journal is respected, or that any meaningful peer review occurs...; cf. Drmies above who has contacted one of the hosting websites: One of the librarians listed on one of the pages you linked confirmed to me what academics already know: a note on a library guide on a university library's website should NOT be taken as any kind of official endorsement for the reliability of that database.

Regarding more in detail the sources you (and Foorgood) have listed, I repeat once again: #1 is a book dedicated to a particular Protestant Christian view and project which is ultimately the same one of Gordon-Conwell; #2 is written by one of the compilers of the ARDA itself (Finke of Pennsylvania University); #3 simply lists and comments the ARDA among other resources, and, note it well, goes back to 2011 when ARDA had not yet switched to the WRD/WCD (it says that at that time its sources were mainly the World Value Survey and the International Social Survey Program); #4 is just the list of references used within the book; #5 is not a source of a good quality and in any case I don't find any reference to the ARDA. Æo (talk) 10:02, 2 November 2022 (UTC)

P.S. O.T. Regarding atheism/irreligion and WIN-Gallup, which are not into question in the present discussion, notice that the definition of "atheism" and "irreligion" can vary according to the context, and that these categories overlap and are not as well definible as belief in a specific Abrahamic religion. Also notice that "atheism" and "irreligion" can overlap with the categories of Eastern religions: Buddhists could be considered atheists, while the notion of "religion" in East Asia does not traditionally apply to non-Abrahamic religions (or to forms of East Asian religions which have adopted an organisational form similar to that of Abrahamic religions), especially to East Asian diffused traditions of worship of gods and ancestors, and the same could be said for certain non-Abrahamic religions and unorganised beliefs which are emerging in the Western world. Æo (talk) 11:25, 2 November 2022 (UTC)

Part 2

Worth observing the Foorgood has repeatedly canvassed editors to this discussion; that's how Ramos1990 got here, for example. --66.44.22.126 (talk) 12:13, 2 November 2022 (UTC)

What are you talking about? Nowhere did he/her or anyone else reach out to me prior to me making a comment on this thread.

Back to the program, to the comments of AEO, I see that most of the discussion is not on whether it is academic (clearly it is - has peer review, Brill is an academic published with peer review, Boston University, notable academics head the project, many peer reviewed publications widely use it, etc). On this alone it is a RS per wikipedia. The stuff about WP:Scholarship applies to fringe publishers, but not Brill (which uses peer review).

Most of the issues that mentioned in this thread relate to methodology and the papers cited in the top of the thread (Liedhegener (2013) and Hsu (2008) both show generally positive views of WRD despite any shortcomings). This has nothing to do with the fact that it IS an academic source, is used by academics to advance research and that it used as a tool in academic research on religion worldwide. As far as I have seen, no major objections have been provided on this latter front. If bias or fault is perceived (this is not agreed upon and the uses of it in peer reviewed publications show its wide utility), this would not be a problem for wikipedia either because even

WP:PARTISAN

states that sources do not need to be neutral and that these sources may be better sources for numerous contexts and that attribution would be appropriate.

On top of that I see that researchers on Islam use it too The Oxford Handbook of Politics in Muslim Societies, and apparently Pew uses WRD data for some of its numbers in Africa for instance The Palgrave Handbook of Islam in Africa. (interestingly it mentions that census data have design problems as well so no dataset is without its problems - which is true since censuses have inconsistent terminology and metrics on religion and some censuses like the American one do not ask about religion). Also, the Liedhegener (2013) paper mentioned at the top of this thread says WCD was used for Encyclopedia Britannica numbers too. Here is another paper using WRD in combination with other sources to get a comprehensive demography [48]. Here is another on Islam in combination with other studies [49]. Also here is one that compares WRD numbers in New Zealand on the nonreligious along with other datasets and is comparable to Pew. Annual Review of the Sociology of Religion: Volume 7

The more I look into this the more I find that it is used quite a bit in the literature along with other datasets either As Is or in a supplementary fashion. I don't see it being used in a depreciative fashion. I see no issue with it being cited with attribution (most studies get attributed either way) and scholars generally do not have a problem with it either (which is why they use it in the first place including general positive comments on it from Liedhegener (2013) and Hsu (2008) which were cited at the top o the thread), and there certainly is no wikipedia policy basis against WRD.Ramos1990 (talk) 22:52, 2 November 2022 (UTC)

"...As far as I have seen, no major objections have been provided on this latter front". What needed to be said has been said, about the use of the source in certain books and about the publishers of either the source itself or books which have made use of it, and academic assessments regarding its non-neutrality and dubious methodology (with one such assessments even affirming that they "seem to have constructed their estimates... from surveys of denominations and missionaries") have been provided (please see

WP:IDNHT

).

Moving forward, please notice that there are various precedents of sources sponsored by or affiliated to religious organisations which have been deemed unreliable:

Gordon-Conwell Theological Seminary. Æo (talk

) 15:10, 3 November 2022 (UTC)

I think we have brought up our points on the matter and when you look at the sources, they show more than what you constantly present in your quotes. The sources you presented here usually support WRD too. For instance, its own methodology paper you cited clearly shows that WRD's sources include Censuses, surveys and polls along with denominational data (see p. 4-5) methodology paper. So just isolating "from surveys of denominations and missionaries" is incorrect on methodological grounds and even this quote misrepresents the source you extracted this from (Woodberry, Robert D. (2010). "World Religion Database: Impressive - but Improvable") because after reading it, Woodberry is very positive to WRD overall and acknowledges its comparativeness with other datasets and even says "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." Plus the fact that WRD is used by Islamic researchers, nonreligious researchers, Pew Research Center (actually integrates it as part of Pew's methodological design per Barton in Palgrave Handbook), CIA estimates (per Woodberry), and Encyclopedia Britannica (per Liedhegener) show that it much more reliable and trusted than you are willing to give credit. But since all datasets have their problems - including censuses, attribution would solve any issues. And its academic status with Brill [50] which is a peer reviewed publisher helps too.Ramos1990 (talk) 00:52, 4 November 2022 (UTC)

Another edition of WRD/WCD data has been the Atlas of Global Christianity (produced by the same Gordon-Conwell team). I have found negative critical assessments even for this edition, this time coming from an "insider" (Christian missionary) source, even though through an academic publisher, written by Anne-Marie Kool of the Evangelical Theological Seminary of Osijek, Croatia:

• Kool, Anne-Marie (2016). "Revisiting Mission in, to and from Europe through Contemporary Image Formation" (PDF). In Charles E. Van Engen (ed.). The State of Missiology Today: Global Innovations in Christian Witness. Downers Grove: IVP Academic. pp. 231–49.
  • p. 1: 《... [the resource] seeks to give “as nuanced a picture as possible” of the history of Christianity over the last 100 years showing an “unmistakable” general pattern, that Christianity experiences a “severe recession” on the European continent that once was its primary base, while it has undergone “unprecedented growth and expansion” in the other parts of the world.》
  • p. 2: 《... widespread caution is raised with regard to the accuracy of the figures and not to engage in statistical analysis with the data, “without robustness checking… they contain random error and probably some systematic error” ...》
  • p. 9, containing a self-criticism from Kool for having herself made uncritical use of the data: 《The World Christian Database and the World Religion Database serve as sources for the data of the Atlas. With regard to the methodology used, Woodberry is right in emphasizing that “more transparency is needed”. 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. 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.》
  • p. 12, about systematic overestimation of Christianity in Europe, with allusions that there might be financial reason behind such overestimations: 《The statistical image of Europe that is now communicated only re-enforces the image of Europe as a Christian continent, by not giving insight in the internal diversification and erosion. So why is only this broad definition used? Is it for fear of losing power? Or for maintaining the image of the numerically strong “World C” that provides the human and financial resources to “finish the task”? Are matters of Christian finance playing a role? Out of a sense of empire building? Or of a sense of hidden resistance to accept that Europe also is now also a mission field? Is it out of fear of becoming a minority? Fear for ending up statistically weaker than the Muslims? Or an attempt to cling to the influence of the “Western” over the “non-Western” world, based on an image of Europe as still a massive Christian continent?》
  • p. 13; it is a missionary tool designed for a specific strategy of aggression towards what in American missionary Christianity has been conceptualised as the "10/40 window": 《Eric Friede’s sharp analysis points us to the fact that the Atlas is ultimately written from the perspective of the so-called Great Commission Christians, Christians who engage in and support Christian missions, as many essays address the issue of “how to grow Christianity” in a particular region. The mission strategy invoked is then one of identifying within Global Christianity the resources needed for the task, the human resources, the GCC Christians, as well as the Christian finances that could make this enterprise work. An assessment of major tools needed for finishing this task is offered in subsequent sections, like Bible translation is followed by a section on Evangelization, with a division of the world in A, B and C, according to the level “being evangelized”. Statistics are used to motivate missionaries and national workers to mission action with Christian mission being reduced to a manageable enterprise with a dominant quantitative approach and a well-defined pragmatic orientation, “as a typical school of thought coming from modern United States”.》 Kool makes largely reference to: Eric Friede, "Book Review. Atlas of Global Christianity: 1910-2010, by Johnson, T.M. & Ross, K.R. (Eds.), 2010", Edinburgh, Edinburgh University Press, Theological Librarianship 3(1), 2010.

Among other sources, some of which we have already analysed, Kool makes reference to:

• Brierley, Peter. (2010). "World Religion Database: Detail Beyond Belief!". International Bulletin of Missionary Research. 34 (1): 18–20.
  • I can't access the paper at the moment. However, it is a critical assessment, once again coming from a missionary journal, that raises doubts as to the reliability of the WRD on the basis of the mismatch of the latter's data (purportedly based on censuses) with actual data from censuses, in particular those of the UK. Judd Birdsall and Lori Beaman, in Faith in Numbers: Can we Trust Quantitative Data on Religious Affiliation and Religious Freedom?, Transatlantic Policy Network on Religion & Diplomacy, 22 June 2020, at p. 3 say that the WRD, despite being widely cited and impressive, "comes with limitations. In his review of the Database, the statistician Peter Brierley pointed out that for the United Kingdom the Database used denominational reports, such as Church of England baptismal records, rather than the UK census figures to calculate affiliation. A tally of denominational reporting showed that 82% of Britons were Christian, whereas only 72% of them claimed to be Christian in the UK census". (n.b. Brierley makes reference to the UK 2001 census data, showing that already in 2001 the WRD overestimated Christianity in the UK by at least 10%).

Æo (talk) 22:13, 4 November 2022 (UTC)

Appreciate the references. However, again these do not impact WRD much at all. The Kool source on the Atlas mentions that WRD data is used by the Atlas, not that WRD is the Atlas or that the goal of the Atlas was the same as WRD. Just like Pew uses WRD, it makes no sense to associate Pew's agendas with WRD's database just because one uses the other. The Atlas' goal and interpretations numerical data is different than WRD and Kool rightly focuses on the contents of the Atlas, instead of WRD (WRD is not mentioned much throughout the paper). WRD is one data set and is not the only one and attribution solves any issues here on wikipedia. You keep on thinking that these papers are calling for the removal of WRD when they are actually trying to improve it and they continuously praise it overall. They all agree that it is very valuable and merely say that there are limitations to it - just as the same applies to Pew, Gallup, and censuses all around the world. This is nothing new in this and if you ever look at the numbers of nonreligion, for example; from Pew, Gallup, and Cenuses, there are significant discrepancies to be found there between these datasets (easy examples include China, Japan, and numerous countries in Europe like Netherlands and Sweden). They are all flawed and limited. None stands as the authority. Stuff like "spiritual but not religious" messes up the numbers because religion is not perceived in Western sense in most of the world.

If we want to criticize census data, there are papers showing the limitations and issues with that too [51]. Censuses get quantity - but they do not guarantee quality of course. In fact Pew's methodology mentions the limitations of surveys and census too (America has not asked on religion for 70 years, for example). Pew clearly states that "Censuses and nationally representative surveys can provide valid and reliable measures of religious landscapes when they are conducted following the best practices of social science research. Valid measurement in censuses and surveys also requires that respondents are free to provide information without fear of negative governmental or social consequences. However, variation in methods among censuses and surveys (including sampling, question wording, response categories and period of data collection) can lead to variation in results. Social, cultural or political factors also may affect how answers to census and survey questions are provided and recorded." Its pretty obvious that big variations exist between just these datasets alone. Anyways, Pew also mention that they used WRD data for 57 countries as a supplement in their methodology in that same section. Furthermore, Pew acknowledges that statistical reports from religious groups are also valid measures. "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."

So I don't see much of an issue in light of this. So all of this thread on equating a critique of a dataset = bad dataset is preposterous when you see that all datasets have problems and issues. In fact, there is research indicating that "religion" is invented in surveys and polls (if you are interested see Wuthnow, Robert (2015). Inventing American Religion : Polls, Surveys, and the Tenuous Quest for a Nation's Faith. Oxford University Press. ISBN 9780190258900.) We know that WRD is used by Pew, CIA, Encyclopedia Britannica, Islamic researchers, and nonreligious researchers among many others. But most importantly it satisfies wikipedia's RS criteria since it has peer review, and is from Brill, an academic publisher.

You cited Brierley, Peter. (2010). "World Religion Database: Detail Beyond Belief!". International Bulletin of Missionary Research. 34 (1): 18–20.. Cool. It says the same thing - that there are limitations. But keep in mind what he also clearly states "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." And after reading it, the overall view is positive to WRD, not negative on WRD.Ramos1990 (talk) 02:21, 5 November 2022 (UTC)

WP:IDNHT

:

• You write: "...these do not impact WRD much at all". This is your opinion, not that of Kool, and not my opinion or that of other editors for whom the WRD's reliability is questionable.
• "The Kool source on the Atlas mentions that WRD data is used by the Atlas, not that WRD is the Atlas or that the goal of the Atlas was the same as WRD". Kool is clear: "The World Christian Database and the World Religion Database serve as sources for the data of the Atlas", and the Atlas is produced by the very same Gordon-Conwell team.
• As for the rest of your message, it is completely off-topic and diverts from the main theme. Censuses, Pew, CIA, Britannica (of which the latter two are not statistical organisations and only cite figures taken from other sources) are not what is being discussed here. Indeed, Pew-Templeton's Global Religious Futures is affected by the present discussion since it, just like ARDA/WCD/WRD, is a violation of
  WP:CRYSTAL
  (pseudodata projections based on Pew's 2001-2010 cycles of surveys, which are being presented throughout Wikipedia as hard data for 2020, 2030, 2040, 2050). The fact that Pew used the WRD for some of its data only detracts from Pew's own quality.

Æo (talk) 14:24, 5 November 2022 (UTC)

The Atlas uses WRD as a source for its data, like you mentioned. So does Pew, Muslim researchers, nonreligion researchers, etc. None of this means that WRD is equivalent to the interpretations, contents, views or arguments presented in the Atlas, Pew, Muslim, nonreligion sources. It is just a dataset. In terms of projections, I am sure you already know that ALL projections are wrong. In the last quarter century it was projected by numerous sources and studies that significant parts of the world would not be religious, due to secularization. But this never happened. Projections are usually wrong. But that is a different discussion.Ramos1990 (talk) 16:57, 5 November 2022 (UTC)

Not being sure, whether it is the right place, I am repeating my comment regaridng sources for membership data in religious bodies. They should be used according this priority ranking: 1. Data of the religious body itself if officially counted like in Austria and Germany 2. Census data like in India, Indonesia and many other contries 3. Data from high quality independent surveys like in Spain or US 4. CIA data may fill the remainig gaps 5. Data from missionary sources should be avoided! Nillurcheier (talk) 10:55, 7 November 2022 (UTC)

Thank you, Nillurcheier, in the next days I will proceed with drawing up a summary rationale and a RfC for the deprecation of missionary and projected data. Æo (talk) 01:07, 12 November 2022 (UTC)

Summary of general consensus

According to authoritative assessments provided hereinabove:

• The following ARDA/WCD/WRD alias
  WP:CRYSTAL
  , since their data are projections from the past (1900) to the future (2050) made on the basis of unclear and non-empirical methodologies (i.e. they lack concrete grounding in data provided by actual surveys of populations):
  • https://www.gordonconwell.edu/
  • https://worldreligiondatabase.org/
  • https://worldchristiandatabase.org/
  • https://www.thearda.com/
• The Pew-Templeton Global Religious Futures dataset (n.b. not other studies made by the Pew Research Center) presents problems similar to those of ARDA/WCD/WRD/GC datasets as to the reliability of projections (
  WP:CRYSTAL
  ; from 2010 to 2050, and sometimes to 2070 and 2100), although in this case the methodology is somewhat clearer (their projections are based on censuses of the early 2000s and Pew's own surveys conducted prior to 2010, the year from which the GRF data were projected; however, note well that Pew's surveys conducted after 2010, focused on religion demographics of the United States in 2014, 2019 and 2021, have already disproven previous GRF projections):
  • http://www.globalreligiousfutures.org/
  • https://www.pewresearch.org/religion/2015/04/02/religious-projection-table/
  • https://www.pewresearch.org/religion/2015/04/02/religious-projections-2010-2050/

N.b. the links may not be exhaustive, as the data have been replicated on various other websites.

The problem with these datasets has arisen since they have been passed off in various Wikipedia articles as hard data (which they are not, as they are projections). In any case, there is general consensus, both in the discussion hereinabove and in past discussions on other Wikipedia talk pages, that projected data and data produced by missionary organisations should never be used to replace census data and data produced by professional (non-partisan) statistical organisations.--Æo (talk) 22:21, 15 November 2022 (UTC)

Correction to the "summary of general consensus"

Parts 1 and 2 of the whole discussion clearly show massive disagreements among editors on everything mentioned in the "Summary of general consensus". There was no consensus.

• The sources used "authoritative assessments provided hereinabove" generally support the use of either of the 3 databases they discuss (ARDA, WCD, WRD). These sources do not depreciate theses databases and some specifically mention that they use these databases themselves. I extracted more quotes at the bottom of some of these sources to show context since it looks like ignoring the positive things they say would distort their assessment. Context matters and transparency is important.

• Pew-Templeton Global Religious Futures was not discussed much either and even here there were clear disagreements among editors.

• The notion of using projection data was not discussed much throughout this whole discussion and in the little that was mentioned, there were clearly disagreements among editors.Ramos1990 (talk) 01:13, 17 November 2022 (UTC)

Replying to your points,

• 1) There is a general consensus that the sources are questionable, both from the present discussion and past ones. In the present discussion, 5 users have raised concerns, while 2 (i.e. you and Foorgood) have continued to repeat the same things to support the datasets. The consensus, from both the present and past discussions, is that these sources should never replace censuses and surveys from statistical organisations; it is not about deprecation, as we have not reached that point yet, and it will require a RfC.

• 2&3) About GRF and projections, I mainly referred to past discussions. In any case, they are against
  WP:CRYSTAL
  and passing them off as actual data is very simply an untruth.

--Æo (talk) 00:49, 17 November 2022 (UTC)

We should leave it up to an uninvolved editor to summarize this whole discussion, not you or me. Other editors have not agreed to your points. Most just provided very few comments early on on what makes the source relaible or not per wikipedia criteria. None of which look like they mention your points one methodology at all. And none said these were unreliable explicitly either.Ramos1990 (talk) 01:13, 17 November 2022 (UTC)

Ramos1990, I agree about an uninvolved closure. However, my summary was not meant to close the discussion, it was meant to provide a rationale for a RfC as the next step. In any case, I have taken part in various discussions concerning this topic over the years, in many of which Iryna Harpy also participated and supported my views; my idea of consensus is built on those past discussions too.--Æo (talk) 01:28, 17 November 2022 (UTC)

P.S. The issue about methodology was first raised by Erp, then further investigated by me, and is mentioned in particular in one of the excerpts from Anne-Marie Kool's paper.--Æo (talk) 01:59, 17 November 2022 (UTC)

RfC: deprecation of Gordon-Conwell's WRD/WCD/ARDA & Pew-Templeton's GRF

Is this source reliable? They appear to have an editorial team. — VORTEX³⁴²⁷ (Talk!) 23:56, 12 January 2023 (UTC)

I don't think so. Having an editorial staff is necessary, but not sufficient. I don't see any other reliable sources citing this website. I note that Rotten Tomatoes does not accept its reviews except where written by a couple of specific reviewers, which means those reviewers are notable and reliable, but not the website itself. If it's not good enough for Rotten Tomatoes, it's not good enough for Wikipedia.

Banks Irk (talk

) 00:41, 13 January 2023 (UTC)

Rotten Tomatoes accepts all kinds of self-published blogs as "Tomatometer approved". Being accepted by Rotten Tomatoes is meaningless. NinjaRobotPirate (talk) 17:56, 15 January 2023 (UTC)

Being rejected by Rotten Tomatoes, on the other hand, is pretty much the kiss of death when it comes to reliability.

Banks Irk (talk

) 19:18, 15 January 2023 (UTC)

Is https://web.archive.org/web/20180321134612/http://georgiatoday.ge/news/3112/American-Ex-Paratrooper-Joins-Georgian-Legion-Fighting-in-Ukraine a reliable source — Preceding unsigned comment added by Ola Tønningsberg (talk • contribs)

I would say yes. It seems like a standard English-language newspaper with an editorial team in a democratic country. Unless there's some disqualifying factor that I'm not aware of, it should be fine. Thebiguglyalien (talk) 20:12, 15 January 2023 (UTC)

A small group of editors at the

• From
  WP:RSP; "For politics and science, there is consensus that the reliability of Fox News is unclear and that additional considerations apply to its use. As a result, Fox News is considered marginally reliable and generally does not qualify as a "high-quality source" for the purpose of substantiating exceptional claims in these topic areas.". Black Kite (talk)
  19:12, 11 January 2023 (UTC)

It's not a "small group of editors" on one talk page, but all of the editors who discussed the issue here. – Muboshgu (talk) 19:16, 11 January 2023 (UTC)

Not all who discussed there agreed. Peter Gulutzan (talk) 14:46, 12 January 2023 (UTC)

Unanimity is not a synonym for consensus... --Jayron32 20:03, 12 January 2023 (UTC)

... which is why people shouldn't claim it. Peter Gulutzan (talk) 15:31, 13 January 2023 (UTC)

The point being that OP is mischaracterizing the consensus as among only a small group of editors. Andre🚐 15:47, 13 January 2023 (UTC)

While Fox is certainly a "main" media outlet, after Trump it has become a fringe media outlet pushing conspiracy theories, fake news, and directly repeating Russian propaganda from Russian media sources. Main does not equal mainstream. At Wikipedia, a source's popularity with a segment of society (increasingly radicalized and far-right wing) does not guarantee reliability. -- Valjean (talk) (PING me) 21:48, 11 January 2023 (UTC)

Fox News should not be considered reliable for this instance. A recent RFC found that it is not generally reliable for politics, and should be used with care for controversial topics.

WP:FOXNEWS recent RFC Andre🚐

22:07, 11 January 2023 (UTC)

The RfC found no consensus on the reliability of Fox for politics. Editors who claim it was found to be generally unreliable/not reliable are misinforming others. I will note Andrevan is correctly stating it was not found to be "generally reliable" but the yellow category is a considerations apply, not "do not use". This is certainly a story where information reported by Fox should not be dismissed on a procedural basis. Springee (talk) 13:43, 13 January 2023 (UTC)

It is "do not use" for controversial claims, which Twitter Files are. Andre🚐 14:24, 13 January 2023 (UTC)

No, it is case by case. Also the Twitter files as a whole are not a controversial claim. You would have to look at the specific claim being added to the article to say "exclude as controversial". Springee (talk) 14:33, 13 January 2023 (UTC)

No, it is not case by case for controversial claims. It is only usable in routine and uncontroversial situations and better sources should be used when available. Andre🚐 14:50, 13 January 2023 (UTC)

I'm glad we agree, case by case for controversial claims. That also means we agree it can be used for controversial claims in some cases. Springee (talk) 14:52, 13 January 2023 (UTC)

Huh? I just said, no it cannot be used for controversial claims. Andre🚐 14:54, 13 January 2023 (UTC)

Well, then you are wrong per the guidelines. "Other considerations apply" != "can not be used". Springee (talk) 15:01, 13 January 2023 (UTC)

No, you need to read it, it is not a high-quality source for controversial claims. It may be used for uncontroversial and routine claims. Andre🚐 15:38, 13 January 2023 (UTC)

• Per the above, this has been discussed many times (see
  WP:FOXNEWS for a list of them), and the general consensus is that while Fox News is generally reliable for some subjects, it should NOT be considered a reliable source for politics, especially for exceptional claims which are not corroborated by other sources. It should generally be avoided; if it is the only such source, then it shouldn't be used, and if its reporting is corroborated by strictly reliable sources, use those instead. --Jayron32
  14:26, 12 January 2023 (UTC)
• Sigh… Has anything changed since the last 10 times we discussed Fox News? If not, give it a rest. Blueboar (talk) 16:07, 12 January 2023 (UTC)

• No it seems not. Slatersteven (talk) 16:18, 12 January 2023 (UTC)
• To be fair, not everyone that comes asking was privy to the earlier discussions, nor knows how to find them. Much better to be empathetic with people new to the area, and gently help them understand the current consensus and where to read the discussions to know how how we arrived there. It may be the 10th time you've dealt with the matter, but it's the OPs first. And when it's your 11th, it will be the NEXT person's first as well. --Jayron32 20:02, 12 January 2023 (UTC)

• Since the RfC was closed as Yellow for Fox News and politics, it shouldn't be excluded as a view on the subject. A number of editors have incorrectly taken the RfC to mean Fox should never be used for political claims other than in about self type instances. That is not what the closing said. Instead it should be evaluated on a case by case basis and typically attributed. If Fox is saying something that may be different from other sources then including with attribution should be fine. Note this does not apply to cases where the Fox News story is referencing a Fox commentator. Springee (talk) 13:46, 13 January 2023 (UTC)

  I agree that it can be used with attribution, though I would try to find something better if possible. Selfstudier (talk) 13:49, 13 January 2023 (UTC)

  Recommend you read the close again as well as the description of a marginally reliable source. The close clearly found a consensus that Fox News is not generally reliable for politics and should not be used for contentious or controversial claims. It may be attributed for opinion, but it should be avoided for situations where it is saying something different than other sources. This RfC establishes Fox News's marginal reliability not through an absence of consensus as to reliability (as determined in the July 2020 RfC) but instead through an affirmative community consensus that Fox News does not qualify as "generally reliable". Andre🚐 14:23, 13 January 2023 (UTC)

  Not green != generally unreliable. Selfstudier (talk) 14:37, 13 January 2023 (UTC)

  Exactly. In this case the specific claims being attributed to Fox would matter. Springee (talk) 14:41, 13 January 2023 (UTC)

  The claims would have to be routine and uncontroversial facts. Andre🚐 14:50, 13 January 2023 (UTC)

  No, if it isn't routine/uncontroversial then it should be reviewed and discussed. It should not be dismissed out of hand based on the false idea that RSP yellow=never use for anything other than.... Springee (talk) 15:03, 13 January 2023 (UTC)

  That's not what is going on. What's going on is that a consensus of editors on the Twitter Files have determined it shouldn't be used for those, and some editors don't understand how the system works. Andre🚐 15:39, 13 January 2023 (UTC)

  Looking at the article talk page the question about Fox wasn't asked well since it just said "Why no Fox?" That could be as simple as "we haven't found a claim that we needed to source to them". However, the answers that amount to a blanket ban on the use are wrong and do not align with the closings of the RSN discussions. Springee (talk) 16:15, 13 January 2023 (UTC)

  There's nothing blanket about it, it's a specific context of controversial political article topics and statements. Andre🚐 18:17, 13 January 2023 (UTC)

  Not "generally reliable" is not the same as do not use for politics. Again, this is the problem with editors who have taken reasonable claims (Option 2, use with care/considerations apply) and turn it into (don't use for politics). With something like the WSJ we generally will assume the information is good if reported there. With Fox we should look and think before including. That does not mean, never use or only use for very basic claims. If Fox is saying something different we need to ask what is being said and in what way is it different? Is it a conflicting interpretation (ie both claims are interpretations but Fox and CNN reach different interpretations)? In that case we could include Fox with attribution. Is it a medical related claim that conflicts with the CDC? In that case we shouldn't include it. Is it a non-interpretive fact that other source having mentions, in that case we can include it assuming it's relevant. That we are cautious when using Fox is a good thing (and should apply to more sources). That we think total exclusion is a good idea moves us towards systemic bias. Springee (talk) 14:40, 13 January 2023 (UTC)

  Fox News generally does not qualify as a high-quality source[] for the purposes of substantiating exceptional claims. That includes basically everything about the Twitter Files and the Covid vaccine claims being made therein. Andre🚐 14:47, 13 January 2023 (UTC)

  Generally dose not!= do not use. Additionally, we need to be clear what counts as an exceptional claim. "Biden was found to be on Russian paylist" = exceptional claim. Without knowing the specific Twitter claim in question we can't say it was or was not "exceptional". Springee (talk) 14:50, 13 January 2023 (UTC)

  There's a consensus of editors on the Twitter Files talk page that it shouldn't be used for the Twitter Files. Andre🚐 14:51, 13 January 2023 (UTC)

  It appears that "consensus" is based on a limited number of editors who are misapplying WP:RSP. I would encourage any editor who thinks they are being stonewalled by editors who assume Fox cannot be used for politics vs should be used with care to start a RfC. Then editors can try to make more coherent arguments and a a RfC closing can decide if "Fox not OK for politics" arguments can be dismissed. Springee (talk) 14:56, 13 January 2023 (UTC)

  No, again, look at both that and this discussion, there is a consensus that Fox News should not be used for controversial claims. A new RFC is certainly not necessary as there is a consensus. Your opinion is not changing the consensus. Andre🚐 14:57, 13 January 2023 (UTC)

  You are wrong about what RSP yellow actually means. Yes, that is how many use it and often numbers trump the actual guidelines in question but that is not the correct read. Additionally, a local RfC on the content supersedes the RSP RfC. It both satisfies the "considerations apply" part and would allow editors to decide if the specific claim in question is actually "contentious"/controversial. Springee (talk) 15:07, 13 January 2023 (UTC)

  By virtue of editors determining this material is contentious and controversial and by determining that Fox News shouldn't be used in this instance, the RFC is satisfied as well as the local consensus. There is no consensus that Fox News may be used in this instance. There is a consensus that Fox News is not generally reliable. Andre🚐 15:40, 13 January 2023 (UTC)

  Since no specific claim was proposed it is inappropriate to issue a general statement that Fox cannot be used. We also do need to be careful when local consensus is based on a relatively small number of editors with a similar view on the subject. I've seen cases where local consensus says one thing but when a broader group looks at the question the consensus flips. Still, since we don't have a specific claim and source in question all we can reasonably say is, there is no blanket ban. Specific claims need to be reviewed on a case by case basis. Springee (talk) 16:19, 13 January 2023 (UTC)

  The local consensus and the project consensus and the noticeboard consensus all say that Fox News is not reliable for controversial political issues. Andre🚐 18:17, 13 January 2023 (UTC)

  I don't think one can make a blanket statement like basically everything about the Twitter Files and the Covid vaccine claims being made therein, it depends. Selfstudier (talk) 14:50, 13 January 2023 (UTC)

  One can make such a statement if one has the support of consensus. Andre🚐 14:51, 13 January 2023 (UTC)

  Sure but we are talking about the RFC here, not a local consensus. Selfstudier (talk) 14:52, 13 January 2023 (UTC)

  But that also means editors who simply argue "Fox not allowed for politics" aren't offering a meaningful reason to decline the information. They might as well say "Fox not OK because we don't know what the fox says". Springee (talk) 14:53, 13 January 2023 (UTC)

  Who is arguing that? Would you read the OP's post again? Fox News cannot and should not be used per the OP's argument. Andre🚐 14:55, 13 January 2023 (UTC)

  The OP asked if Fox is blacklisted. The obvious answer is no it is not. Since the OP didn't say what specific claim was being added we can't answer if it considerations have applied or not. Springee (talk) 14:57, 13 January 2023 (UTC)

  The OP is referring to a conversation about the Twitter Files. Andre🚐 14:59, 13 January 2023 (UTC)

  The question here was if Fox can be used at all, no specific claim is or source article has been presented here. Springee (talk) 15:00, 13 January 2023 (UTC)

  You just need to read what the OP wrote again, because that is not what the OP wrote. Andre🚐 15:42, 13 January 2023 (UTC)

  That's not clear to me, the OP should refile like the old days, is Fox News reliable, with attribution, for (some statement). Selfstudier (talk) 14:57, 13 January 2023 (UTC)

  That's a good suggestion. One of the worst things to happen at RSN was the creation of RSP. It basically encourages editors to try to get sources bucketed based on their preference rather than doing what RS says we should do, consider if the source is reliable for the specific claim in question. Springee (talk) 14:59, 13 January 2023 (UTC)

  The ship has sailed on this. Andre🚐 15:41, 13 January 2023 (UTC)

  Let's see if the OP does refile, I am withholding judgement until then. Selfstudier (talk) 15:50, 13 January 2023 (UTC)

  Not quite. One of the key questions with marginal sources (which is what we concluded Fox is, when it comes to politics) is whether a claim is

  WP:EXCEPTIONAL or not. It can be used for simple unexceptional statements; it can't be used for shocking things, political scandals, or other claims of an exceptional nature. This is article is clearly focused on the latter - it posits an extensive conspiracy between various political actors, the government, and the former management of Twitter based on virtually no evidence or on evidence that essentially all reliable sources have said they're misrepresenting. More generally, it should be obvious that when it comes to this topic, statements that are only citable to Fox are quite frequently going to be exceptional - a common problem when trying to cite marginally reliable sources, so it generally shouldn't be used here. Fox is still used in all sorts of places for all sorts of unexceptional claims, but the fact is that most of the areas we're likely to come into dispute on are going to be cases where Fox says exceptional things. (In fact, OP essentially acknowledges that this is exceptional when they say that only Fox is writing about these aspects - the easiest way to demonstrate that something is unexceptional is to show multiple sources covering it.) More generally, the simple reality is that yellow sources should be replaced by higher-quality sources whenever possible; and when they can't be replaced because they're the only source for something, that's often a reason to examine that statement very carefully to make sure it's not exceptional. --Aquillion (talk

  ) 22:43, 15 January 2023 (UTC)

I've read the Fox News articles on the Twitter Files in comparison to more typical media, and the usual problems with Fox and politics shine there - they are taking the claims of Elon Musk and those he shared the files with as undisputable facts, whereas most media point out that the files only point to typical moderation choices that would be handled at a large internet forum board. The only place where Fox news has any relevance is they are one of the dogwhistles to try to boost the importance of these files beyond what they actually are (something we can source to more reliable sources) Masem (t) 15:10, 13 January 2023 (UTC)

I would still like to see what it is exactly the OP seeks to have an opinion about, it reads more like a complaint about our classification of Fox and doesn't actually say anything specific about Twitter files (I had to look that up). Selfstudier (talk) 15:17, 13 January 2023 (UTC)

Twitter files is 100% political, as it is claims of Twitter censoring conservative voices including collusion with the federal gov't to do so. Given that most mainstream sources pretty much ignored the antics of the Twitter Files while the right-leaning media (including NYPost and WSJ, in addition to Fox and others) dove into the claims, it tells us that we need very careful handling of how they should be reported, and certainly not taking the original reports at their word. Masem (t) 15:28, 13 January 2023 (UTC)

OK, but if there is attribution, we are not taking the reports at their word, we are not making claims in WP voice as fact. Selfstudier (talk) 15:32, 13 January 2023 (UTC)

The contents of the Twitter Files (which are the original stories posted by individual journalists typically not associated with any work) should definitely be quoted and attributed if used, and attributed if summarized. That said, when we summarize, we should not summarize ourselves from the primary source (that clearly would include OR), so we need a secondary source to tell us what the summary is. That is where we need reliable secondary sources that shouldn't be stretching the truth of the original primary source (eg: we should not interpret Supreme Court rulings ourselves but rely on this type of reporting from secondary sources to tell us what is important, and to that end, we want good, quality RSes for that that will not twist the truth).

Fortunately in the case of the Twitter Files, we do have RSes that explain the take that right-leaning sources are taking, alongside a less nuanced take on what the impact of the files really are, so there's no need to evoke the primary source (outside of showing where they came from) and poor RSes like Fox. Masem (t) 15:46, 13 January 2023 (UTC)

@Masem Wow, thank you for that enlightening and not at all tedious explanation of how to properly attribute sources. I mean, who needs to actually read the primary source when you can just rely on some 'reliable secondary sources' to tell you what's important? And thank goodness we have those trusty 'RSes' to guide us away from those pesky 'right-leaning sources' and protect us from the dangers of actually thinking for ourselves. Truly, this is the height of Wikipedian integrity. Man-at-Bogomil (talk) 05:11, 15 January 2023 (UTC) [troll blocked; Drmies (talk) 16:20, 15 January 2023 (UTC)]

This is a core aspect of

WP:PSTS. We prefer reliable secondary sources over reliable primary sources as the secondary sources will help explain why something is important and avoid editor's own OR to make such a claim (as it would absolutely be for a case like the Twitter Files). Masem (t

) 05:24, 15 January 2023 (UTC)

I think this gets to what I've argued all along, we need to see the specific claim and the specific source. I agree that if "Musk says X" we shouldn't treat that as anything other than a likely self serving claim made by Musk. If Fox quotes it we have to decide if Musk's opinion on the subject is DUE. If Fox treat it as fact we should not and should consider if it should be included (I'm not inclined to believe Musk). However, we should not kneejerk reject any inclusion of Fox as a source. Springee (talk) 16:30, 13 January 2023 (UTC)

Masem is exactly right. Andre🚐 15:41, 13 January 2023 (UTC)

• I don't see where the IP/OP has provided any example here nor there. So, why are we wasting time? O3000, Ret. (talk) 16:38, 13 January 2023 (UTC)
• The nuance of our RSP listing is important… while it is accurate to say that we do not consider Fox generally reliable, that does not mean we consider it generally unreliable. It is currently in a wishy-washy in-between state. And THAT means we have to examine whether it is or is not specifically reliable (ie reliable for a specific statement, in a specific article). So, until we know the specifics, we can not answer the OP’s question, except to say “no, we have not banned Fox outright… but there are restrictions, so we need more information”. Blueboar (talk) 17:24, 13 January 2023 (UTC)

I don't understand why we're ignoring that OP has provided a specific article and it is a controversial one. In that case, no, Fox is not reliable. Andre🚐 18:16, 13 January 2023 (UTC)

The OP only asked one question Has Wikipedia blacklisted Fox News site-wide, or only in political articles frequented by the same established editors and admins? Selfstudier (talk) 18:18, 13 January 2023 (UTC)

Just glossing over the opening sentence that elucidates the context. A small group of editors at the Twitter files article have refused to consider the use of Fox News as a source for the article. Andre🚐 18:21, 13 January 2023 (UTC)

Context, background, who knows? That's why I said I would wait and see if OP refiles, if not there is nothing to debate. Selfstudier (talk) 18:24, 13 January 2023 (UTC)

Yup… a dif or two would be helpful here, so we could see what specifically is not being accepted and the context in which the OP wishes to cite it. Blueboar (talk) 19:42, 13 January 2023 (UTC)

Response to IP. I believe the Wiki-community has already decided on which news outlets are reliable, which aren't & anywhere in between. GoodDay (talk) 00:47, 14 January 2023 (UTC)

• The focus of the article is clearly on exceptional claims about politics, which are not appropriate to cite to a marginally reliable source like Fox. --Aquillion (talk) 22:43, 15 January 2023 (UTC)

See

WP:FORBESCON, Forbes staff and Forbes contributors are two very different sources. _signed, Rosguill ^talk

21:11, 15 January 2023 (UTC)

Thx -

WP:WAWARD

) 21:19, 15 January 2023 (UTC)

The same content regarding her ballet experience is included in this podcast. The aformentioned contested verifiability edit summary says something about this source being OK. Can I add that she studied ballet in her youth based on the podcast content?-

WP:WAWARD

) 21:23, 15 January 2023 (UTC)

I guess that would be a listed exception. Sorry.

Resolved

--

WP:WAWARD

) 05:50, 16 January 2023 (UTC)

I see sometimes on the reliable sources verified accounts on twitter can be a source. I was wondering if the account is not verified but it is shown in multiple reliable sources such as the time, the Guardian... that they mention the username that it's the real one, can it be used then?

For more informations , I found a draft called Draft: Edward Hayter and have been editing it since then as he is one of my favourite actor. The reason of the main question is because he doesn't have much articles about him but he has collaborated with Victoria Emslie and she has stated in these reliable sources her twitter username, and she does mention him and it would make more reliable sources and content I could use such as proving he was in a music video with her

Veganpurplefox, if I understand your question correctly. If the Guardian states "this twitter account belongs to that person", then IMO that "works" for

WP:SPS is quite strict here: "Never use self-published sources as third-party sources about living people, even if the author is an expert, well-known professional researcher, or writer." Emphasis in original. So it will not help you in WP-land. Gråbergs Gråa Sång (talk

) 08:52, 16 January 2023 (UTC)

@

Fram: I've run in to this one a thousand times, always on those kinds of articles. It's a mix of limited original content, and a large amount of aggregate news feeds and paid placement/press release material. The only point in their favor is that they do not try to hide the source; articles have a clear indicator of which press release feed (iNewswire, CDN Newswire, Newsmantraa, TheExpressWire, etc) provided the material, or if it from a "real" news feed. That one is "GetNews", which seems to just spew out PR and crytobro stuff, so not in any way RS. Sam Kuru (talk)

15:20, 14 January 2023 (UTC)

Yep, I ran across another new article using such a "Get News" promo article as a source. Is there an easy way to get this source restricted or blacklisted?

Fram (talk

) 12:16, 16 January 2023 (UTC)

Links to this site [79] are used quite a bit, and I note that when the site does source it is often or usually to conference papers, etc. In the actual article searching for conf. turns up 9 conference papers. Also 10 uses of ironically a Roger Weller [80] who is a geography instructor at Cochise College which offers associate degrees. 15 links to meetings. Most if not all of the two word citations are links to the Russian site. A bit of a mess. Am I correct about my assessment of these sources, esp the Russian one? Thanks. Doug Weller talk 13:26, 16 January 2023 (UTC)

I think there's a question of list inclusion criteria, and whether these references are even required. RSN is for global discussion of specific sources. Let's take the discussion to Talk:List of possible impact structures on Earth. — hike395 (talk) 13:49, 16 January 2023 (UTC)

Is this page at Soap Opera News relaible for the date of birth of actor Mike Manning? Nightscream (talk) 20:19, 16 January 2023 (UTC)

The source doesn't really pass the "sniff test". The About Us page looks rather amateurish. For certain banal things like perhaps episode synopses and air dates and stuff like that, it might be better than nothing, but I wouldn't hang any

WP:BLP-type stuff on it, especially not birthdates. --Jayron32

02:12, 17 January 2023 (UTC)

updating software 223.233.74.137 (talk) 10:11, 17 January 2023 (UTC)

What? Slatersteven (talk) 10:48, 17 January 2023 (UTC)

OhNoitsJamie sent this.

Final warning; please take it to