Wikipedia talk:New page patrol source guide

By Topic

For by topic how will what you're thinking compare to WP:RSP? Best, Barkeep49 (talk) 04:24, 12 March 2019 (UTC)

I'm not sure I understand the question. At any rate, my general idea was that we don't need all of the information and justifications of RSP, but to simply list entries from RSP where most relevant (either by region or topic). There's some ambiguity about where some sources could go (particularly unreliable ones), but I think as long as sources are searchable somewhere in the article the information is conveyed well enough. _signed, Rosguill ^talk 04:32, 12 March 2019 (UTC)

I've now had occasion to reference this a couple times and am glad you're developing it Rosguill. Best, Barkeep49 (talk) 18:00, 21 March 2019 (UTC)

Spark/Cliff Notes

What evidence do we have that they would be considered RS for notability purposes? I'm not seeing it from the RSN discussions. I find it hard to believe that if something is in Sparks/Cliffs it's not notable but it should also be relatively easy to establish that from other more reliable sourcing. Best, Barkeep49 (talk) 17:14, 28 March 2019 (UTC)

Barkeep49, if they're study guides, then by definition any text covered by them is a text that is actively studied and hence notable. I'll admit that it is reading between the lines somewhat, and agree that there will almost certainly be other sources available in the event that Spark or Cliff covers them. _signed, Rosguill ^talk 17:16, 28 March 2019 (UTC)

Rosguill Hmm. The shortcut to determining that they're actively studied (relevant for several SNG) is an interesting argument. Still feels tenuous given the low level of discussion but makes some sense. Best, Barkeep49 (talk) 17:25, 28 March 2019 (UTC)

Barkeep49 btw, I'd appreciate your input on the various points that I raised on the NPP discussion page, as they concern how to further expand this page, and I'm a bit hesitant to start doing major expansions on my own. _signed, Rosguill ^talk 17:27, 28 March 2019 (UTC)

Rosguill Hadn't seen that before - must have gotten buried on my watchlist. Will read and respond soonish. Best, Barkeep49 (talk) 17:28, 28 March 2019 (UTC)

France: Source Reliability

Le Monde, La Liberation and Le Figaro are the French newspapers with the most journalistic integrity (evidence over opinion, citing sources, etc.).

I noticed you discredited Patheos as a source, but some of its 'bloggers' (some are respected journalists) are strict about citing their sources and peer-review... depends on the blog.

Anyhow, thanks for your efforts - got word of them through a newsletter sent to my talk-page. Cheers. TP ✎ ✓ 20:17, 17 May 2019 (UTC)

ThePromenader, Thanks, I agree with your suggestions regarding French sources, and would also add France24 and Le Diplo. I was able to find an RSN discussion to back up the assessment for Le Diplo here [1], as well as passing references to Le Figaro (I couldn't find anything for La Liberation because search results were overwhelmed by "liberation" being used as a word in other contexts). The issue at this point is what level of demonstrated, pre-existing consensus is needed to expand the list. Based on these discussions, I would feel fairly confident adding Le Diplo. I'm tempted to start discussions on RSN, but am unsure whether it would be better to go one-by-one for sources, or to periodically start discussions asking for editors to provide and comment on the general reliability of prominent sources for a given region or topic. _signed, Rosguill ^talk 22:06, 17 May 2019 (UTC)

I think where there's not good consensus bringing your version of sources to RSN as a group is a good starting point. Best, Barkeep49 (talk) 23:40, 17 May 2019 (UTC)

How ready is this?

Have we reached a point where there are enough sources that putting together a Google Custom search engine to go through them and advertising the availability of that tool is viable? Best, Barkeep49 (talk) 17:57, 10 June 2019 (UTC)

I'm honestly not sure. Currently over 90% of the sources are just from RSP, and we're missing coverage from much of the world. That having been said, publicizing this may help fill out those gaps. _signed, Rosguill ^talk 18:12, 10 June 2019 (UTC)

Yeah I'm not sure if that will publicize or give a false sense of it being ready for NPP use. Since a GSE already exists for RS/P (which I love using when doing NPP) I'll continue to wait until we fill in more of the global newspaper coverage which is where I think the strength of this endeavor ultimately lies (and also in sources too obvious to have RS determinations but obviously RS like when I added ESPN for sports). Best, Barkeep49 (talk) 18:32, 10 June 2019 (UTC)

While this is still a work in progress it feels far more complete at this point. Is that right Rosguill? I see it having two uses - first to check sources that exist in a page for reliability in establishing notability and second for finding sources not present which indicate notability. For this second purpose, I have gone ahead and created a Google Custom Search Engine based on the current status of this page - which gave me 217 sites. I am going to try it out and see how it compares to other methods I have when doing NPP; it will pull only English language sources when I search in English but if I use another script (e.g. Arabic) it'll pull results that way too. I figure Onel5969 and JTtheOG might also be good "beta" testers before we say anything to the wider NPP community. In really examining the places I pulled from judgement will still be necessary - the clearest example being Forbes where staff writers are RS but contributors are not. There were a few other sources, around pop culture in particular, who are of limited reliability (or maybe only reliable for staff reviews and not news/user reviews). So it'll be interesting to see how useful, if at all, this proves to be while doing reviewing. Best, Barkeep49 (talk) 21:31, 21 October 2019 (UTC)

Barkeep49, I think the real potential value of this is to have a centralized location with source information for relatively obscure regions (by Wikipedia's standards). 99% of new reviewers are not going to know what sources are reliable for say, Somalia, and this guide can help people find their footing for such cases. As such, the coverage is still very spotty: e.g. we're missing virtually the entirety of Latin America, but we have enough for the Czech Republic to be useful there. So it's usable, even if it's still nowhere near complete.

I think the issue we're running into now is that there isn't really a kosher way to elicit reliability assessments for these underserved regions; RSN RfCs are controversial. Maybe we could get some more acceptance for this case if we make it clear that our intent is to combat systemic bias and that we're generally not trying to deprecate sources through this process (which seems to be most people's objections to RSN RfCs). The alternative would be to have NPP reviewers (and also volunteers from other relevant projects) collaborate on building this list without necessarily taking the discussions to an RfC at RSN, or to keep the status quo of monitoring RSN like a vulture but not doing much to initiate discussions that address our blind spots. _signed, Rosguill ^talk 23:51, 21 October 2019 (UTC)

Yes I agree that the value comes less for the sources we have lots of discussion about and more for sources we don't. I think given the discomfort over RSN RfCs, I think if we could attract 10 (or more) editors who would be willing to consider reliability of a source here that could be good enough for a consensus about how to handle for NPP. The best way to get that - as I think we could draw some from NPP but would want to reach a much broader constituent base - requires some thinking. Best, Barkeep49 (talk) 00:10, 22 October 2019 (UTC)

One natural group of editors to reach out to is the WikiProjects relevant to the specific sources/countries/regions/topics at hand, although depending on the level of activity at that project this may not get us many more participants. _signed, Rosguill ^talk 00:24, 22 October 2019 (UTC)

That is an obvious group but I think many Wikiprojects are inactive and so ultimately aren't helpful. For instance WP:WikiProject Africa is inactive which is too bad because that continent is one where there's a lot of needed work. I'm curious how Wikipedia:Reliable_sources/Noticeboard#RFC:_Moratorium_on_"general_reliability"_RFCs closes. I could see a version where we ping there, ping projects, ping NPP, on some sort of fixed schedule (once a month? once ever six weeks?) for a bunch of sources in a given group (e.g. Vietnameese sources one time, Film the next, and Kenyan the third) and see if that gets enough discussion to decide, for the purposes of giving NPP a clue, enough. That's my half-thought through idea that came to mind first, but more thought is needed there and further thought might produce a better idea anyway. Best, Barkeep49 (talk) 00:48, 22 October 2019 (UTC)

I think that if that discussion is closed as "no moratorium" the plan of regular RfCs to cover blindspots is a great idea. We could probably also schedule some for unrelated regions at the same time (maybe 2 at a time?). Whoever is coordinating a given week's RfC should probably do some basic research on the most popular/influential publications from a given country and present that so that people have an idea of what to talk about. On another note, we should probably use a different framing for these discussions than what has now become the standard for source reliability RfCs. Our goal here isn't to canonize sources as reliable or unreliable, it's to establish a baseline reputation for further assessment, and to distribute the resource of editors' collective experience with sources. _signed, Rosguill ^talk 05:27, 22 October 2019 (UTC)

Another thought would be to try to reach out to other language Wikipedia projects whenever their expertise may be relevant. _signed, Rosguill ^talk 19:03, 22 October 2019 (UTC)

Reaching out xwiki is a good idea. So let's see how that RfC on RfCs gets closed - it's possible that it gets closed with a moratorium but in a way that we could still do something there - just not post to RSN. Or it could be closed with no moratorium. It's a sprawling discussion so I'm not surprised it's remained unclosed for quite a while. Best, Barkeep49 (talk) 20:40, 22 October 2019 (UTC)

It seems like the main reason editors were supporting a moratorium was opposition to the idea of deprecating a source (which many of the RfCs indeed were set up to do), so as long as we frame the discussions as not being about that we may be able to dodge a formal moratorium. That having been said, I'd be rather surprised if the discussion is closed as anything other than "no consensus", which in this case I believe would mean "no consensus to implement a moratorium". _signed, Rosguill ^talk 20:59, 22 October 2019 (UTC)

Well, looks like we got a no consensus, albeit one that included the statement there is a somewhat strong consensus for discouraging RfCs for any source whose reliability has not been previously discussed on RSN or elsewhere as per WP:RFCBEFORE. The inclusion of RFCBEFORE in that statement would imply that this is a caution against running to start an RfC when you just have a single case of wondering if a source is usable in a given article, so we can perhaps make the case that our proposal to do RfCs about sources for various topics and regions comprises a different scenario than the above. However, the fact that we have to make that argument at all may mean that it's best to take this to the village pump proposals board. Although, we might want to workshop this a bit in the ideas board first. _signed, Rosguill ^talk 23:02, 22 October 2019 (UTC)

My thinking is that we work out some of the logistics of how we'd do setup, come up with an example discussion (including sources) so people can wrap their hands around a practical rather than abstract idea, and start with WT:NPR. If that goes well then bringing it to the village pump for wider feedback/acceptance. The challenge here is that RfCs, in my experience, work best when they're focused and so by its very nature discussing 5 - 10 sources (or more) at a time will present challenges. It feels like this idea might be a little firmer in your head than mine. Do you want to throw out a structure for us to start to brainstorm from? Best, Barkeep49 (talk) 14:46, 23 October 2019 (UTC)

Barkeep49, I'm going to draft a sample prompt in User space and then ping you here when it's ready. _signed, Rosguill ^talk 17:58, 23 October 2019 (UTC)

Barkeep49, here's a draft User:Rosguill/TurkeyNPPSGRFC. _signed, Rosguill ^talk 20:35, 23 October 2019 (UTC)

Rosguill, this looks fantastic on the whole. Like you've done top notch work here. My question is - are the summaries for each source helpful or will people feel like it's a finger on the scale? Not even sure it needs to change but throwing that out there. It feels like the concept is ready to take to NPP? Best, Barkeep49 (talk) 22:02, 23 October 2019 (UTC)

Barkeep49, thanks! Regarding the summaries, I think some sort of general introduction is necessary in order to provide context for the discussion, although perhaps that can just be converted to a list of RS and Wikipedia articles that provide overviews. As for the individual subsections, I think that some basic uncontroversial information can be provided (i.e. language of publication, ownership, past RSN discussions), and anything beyond that I'm ok with moving to the Discussion section as a comment attributed to me. I'm realizing that a side effect of putting all this together is that we're probably going to collect enough sources to write (or rewrite) our own articles for these various sources. _signed, Rosguill ^talk 22:06, 23 October 2019 (UTC)

Barkeep49, re bringing this to WT:NPR, did you want to start that discussion? If not I can do it. _signed, Rosguill ^talk 22:29, 23 October 2019 (UTC)

Rosguill, I am happy to do it if you want (please say so) but this is your deal and so I had presumed you would start the discussion. If you want me to, however, just let me know. Best, Barkeep49 (talk) 22:41, 23 October 2019 (UTC)

The discussion at the Village Pump went well. I think we should proceed with the first RfC set. Best, Barkeep49 (talk) 17:14, 20 November 2019 (UTC)

I agree that we should move on to the RfCs. However, I don't think we really got any feedback on the draft discussion beyond general approval for having an RfC, do you think we need to workshop the draft at all or should we move forward with it? _signed, Rosguill ^talk 17:44, 20 November 2019 (UTC)

Honestly? I don't know. I tend to lean in to the BOLD approach but that can definitely backfire in an RfC. Maybe one more ask about draft discussion at NPP? Best, Barkeep49 (talk) 17:52, 20 November 2019 (UTC)

Honestly I think we're fine to go ahead with the existing draft. The only other concerns would be

1. whether we want to wait until either the NPP peer review or Arbitration Committee elections are over (or both) to avoid creating a bureaucratic backlog for interested editors
2. whether we want to reach out to relevant other language wikipedia projects ahead of time in order to be able to field questions better, or if we should just invite them once the discussion has been opened. _signed, Rosguill ^talk 23:08, 20 November 2019 (UTC)

   Rosguill, I think the idea of waiting until after the holidays to do the first of these makes sense timing wise. Are we still thinking Turkey for the first one? If so https://tr.wikipedia.org/wiki/Vikipedi:K%C3%B6y_%C3%A7e%C5%9Fmesi seems like their Village Pump Barkeep49 (talk) 23:31, 20 November 2019 (UTC)

   Barkeep49, by holidays do you mean through the New Year? I'd be fine with that. As far as which region to start with, I think it's fine to do Turkey since we already have it drafted out; going forward we should prioritize regions that we have disproportionate trouble with at NPP (IMO Vietnam, India, Iran, China/Taiwan and Russia are high priority in this sense). I used Turkey as our starting example because I wanted to demonstrate how we would handle sources already mentioned at RSP (in this case, TRT World), and because I'm competent enough at reading Turkish to do some research myself.

   Also, we should probably check the embassy rather than the village pump for languages for which we don't have fluent volunteers (which is true for Turkish, I believe) _signed, Rosguill ^talk 23:40, 20 November 2019 (UTC)

   Rosguill, I was talking after Jan 1. That would be a bad time (Lunar New Year) for some Asian countries. I'm not particularly wedded to Turkey just thought it made sense since you'd done the work. In looking at the NPP Browser I see:

   • 54 results for Ghana (one that I struggle with despite English language sources)
   • 52 for Iran (a good choice IMO)
   • 70 for Russia
   • 8 for Vietnam
   • 27 for Malaysia
   • 26 for Indonesia
   • 338 for India (which I think is too big a country for it to be a good first one but maybe a good second one)
   • 47 for Pakistan
   • 38 for China (which is also a huge country and will be hard to do owing to both size and state media control)
   • 9 for Turkey

   Obviously this is only 1 (impercise) snapshot in time but might help guide us for which countries are truly high need. But it suggests to me that either Ghana (which along with Nigeria are frequently frustrating for me) or Russia (which makes sense) would be highest "need". Best, Barkeep49 (talk) 23:53, 20 November 2019 (UTC)

   Barkeep49, well if we're giving ourselves a month then we should have enough time to draft a discussion for another region. As long as we get to Turkey eventually, the work won't be wasted. The snapshot seems reasonable (a bit surprised to see so much Ghana, maybe those get taken care of earlier in the queue than where I patrol). Maybe we can present these numbers to the rest of NPP and see if anyone has suggestions? On a related note, it may make sense to split India up across a few discussions, focusing on different regions/languages _signed, Rosguill ^talk 00:01, 21 November 2019 (UTC)

   Rosguill, I love the idea of presenting to the broader NPP community. Best, Barkeep49 (talk) 01:49, 21 November 2019 (UTC)

'No consensus'

Rosguill, @Barkeep49: I see lots of sources with no consensus. Fine. However, these really aren't helpful for the common reviewer. Is it worth initiating further discussions at the reliable sources noticeboard to determine whether they are reliable or not? If we don't, then sources could forever sit at 'no consensus' and provide confusion and pointless arguments in the future. Willbb234^Talk (please {{ping}} me in replies) 22:28, 14 January 2020 (UTC)

Great job by the way, it's coming together nicely. Willbb234^Talk (please {{ping}} me in replies)

@Willbb234: Rosguill gets the credit for the good work. I'm just along for the ride. That said I think there is value in listing on conesensus for items - it means the NPP needs to use the skills and expertise they're supposed to have to make a decision for that source in that instance. That said, there is the idea of semi-regular discussions being had to try and narrow that down for given topics. Right now Ghana is scheduled to be the first area looked at. Best, Barkeep49 (talk) 23:31, 14 January 2020 (UTC)

@Willbb234: IMO the value of listing no consensus results is that it provides us an opportunity to list a link to that discussion, which may still be of use to someone trying to consult the list. I'd also go out on a limb and say that in most cases, no consensus means "good enough to count toward GNG" because it indicates the lack of a consensus against using that source. The exception would be if no consensus sources are being used to support a topic that is either very contentious (e.g. a politically convenient murder getting played up by NC sources that stand to benefit from coverage) or very nebulous (e.g. vaguely defined social constructs). _signed, Rosguill ^talk 00:49, 15 January 2020 (UTC)

Initial proposal for Spain

Here's my initial proposal for Spain:

Caution

Contentious topics in Spain with known episodes of controversial coverage or conspiracy theories published in otherwise reliable sources: terrorism (specially 2004 Madrid train bombings and ETA), Catalonia independence movement (specially, 2017 Catalan independence referendum) and the Alcàsser Girls case. Additional caution applies to stories published between 1936 and 1975 (see Francoist Spain) and, to some degree, between 1975 and 1977/1982 (see Spanish transition to democracy).

Generally reliable

• EFE, high reputation news agency, reports published in local press with an EFE byline may be presumed to have higher reliability.
• Europa Press, high reputation news agency, reports published in local press with an EuropaPress/EP byline may be presumed to have higher reliability.
• RTVE, state broadcaster
• El País (publisher: PRISA)
• El Mundo (publisher: Unidad Editorial)
• Público
• eldiario.es
• La Vanguardia (publisher: Godó)
• 20 minutos
• ABC (publisher: Vocento)
• El Periódico de Catalunya (publisher: Prensa Ibérica, main region: Catalonia)
• La Voz de Galicia (main region: Galicia)
• El Correo (publisher: Vocento, main region: Basque Country)
• La Nueva España (publisher: Prensa Ibérica, main region: Asturias)
• Heraldo de Aragón (main region: Aragon)
• Levante-EMV (publisher: Prensa Ibérica, main region: Valencian Community)
• Las Provincias (publisher: Vocento, main region: Valencian Community)

Disputed reliability

• Okdiario, digital newspaper often involved in reliability controversies, including defamation lawsuits. Possibly comparable to Daily Mail in reliability.

Not reliable

• Mediterráneo Digital (not to be confused with El Periódico Mediterráneo), media outlet often originating and spreading hoaxes.
• Caso Aislado, obscure media outlet often originating and spreading hoaxes.
• Contando Estrelas / Elentir, blog often originating and spreading hoaxes.

I have not included bias concerns, since I'm not sure they are relevant to determine notability. For example, El País is generally considered a high quality source, yet they have a history of aggressive editorial line against some governments in Latin America. El Mundo is generally reputable, but had some episodes in the past where it engaged in the spread of conspiracy theories about the 2004 Madrid train bombings. RTVE is generally reliable and with certain independence, yet early coverage of the Iraq War was highly controversial. ABC has a strong right-wing bias, while El Público has a strong left-wing bias. Despite all these considerations, I still think all of them are good for the purpose of notability.

The list of generally reliable sources could be expanded with many more regional newspapers. I included initially those with larger circulation. Some smaller media outlets are included in the not reliable since their usage is high in social media, specially linked to hoaxes, conspiracy theories and other forms of misinformation. --MarioGom (talk) 09:39, 24 April 2020 (UTC)

I have not included official regional TV stations, as they are often the center of political disputes. I think EITB (Basque Country) is generally reliable, and TV3 (Catalonia) used to have a high reputation although it is currently disputed regarding events related to the independence referendum. Telemadrid is quite controversial and Canal Sur (Andalucia) probably too. --MarioGom (talk) 09:53, 24 April 2020 (UTC)

MarioGom, these categorizations seem reasonable to me, but I'm a bit concerned that this proposal is out of process, as this is less a community consensus than you simply proposing additional entries for the list. Even if I or other editors were to wholly endorse the recommendations made here, I'm not sure that this talk page has enough visibility for a local consensus here to be a valid justification for a list entry. _signed, Rosguill ^talk 17:04, 24 April 2020 (UTC)

Rosguill, sure, makes sense. Let's leave this in the discussion page at the moment. --MarioGom (talk) 17:08, 24 April 2020 (UTC)

Publishers

Would it make sense to have a book publishers section? I see Cambridge University Press listed, but no others. I'd suggest

==Publishers==

In many discussions, users clarified that no publisher's works can be considered always reliable for everything.

Generally reliable

• Cambridge University Press
• University of Chicago Press
• Springer Science+Business Media, not to be confused with Axel Springer SE or Springer Publishing
• Greenwood Publishing Group (and subsidiaries)
• HarperCollins Canada

Unreliable

• Pentland Press -- vanity press
• Cambridge Scholars Publishing -- vanity press
• Lulu.com -- self-publishing; deprecated
• Scribd -- self-publishing
• Creative Crayon Publishers

No consensus

• University Presses-- in a discussion, some users felt that they should be considered de facto reliable, while others advocated for a case-by-case basis.
• Jessica Kingsley Publishers-- history of publishing plagiarized content
• New Leaf Publishing Group-- WP:FRINGE applies

References

I haven't looked too extensively, and haven't read through discussions like Wikipedia:Reliable_sources/Noticeboard/Archive_37#One topic publishing house acceptable as RS?. I think a section like this would be useful because it's very hard to tell between a reliable and unreliable publication unless you already know about it. Courtesy ping Rosguill and Barkeep49 Eddie891 ^Talk _Work 20:06, 10 September 2020 (UTC)

Eddie891, I think that this is a good idea. _signed, Rosguill ^talk 20:14, 10 September 2020 (UTC)

Rosguill, I also think that including the above list of publishers is a good idea. I was thinking the same thing as I went through NPPSG today (that it would be good to include some publishers). Seeing as everything in the above list is cited to RSN, I think it is ready to drop into NPPSG soon. –Novem Linguae (talk) 12:59, 6 February 2021 (UTC)

Novem Linguae, I wonder if it makes sense to include them in a subsection for publishers, or if we should try to sort them by their coverage of subjects and/or country of origin? _signed, Rosguill ^talk 13:01, 6 February 2021 (UTC)

My suggestion would be to keep them grouped in a "Publishers" section for now. If it gets bigger and starts including a lot of areas of the world, then it might make sense to start dicing it up. –Novem Linguae (talk) 13:07, 6 February 2021 (UTC)

Seems reasonable to me. _signed, Rosguill ^talk 17:05, 14 February 2021 (UTC)

Globe and Mail

Was recently asked on IRC if The Globe and Mail is an RS. The mentions at RSN are sparse though it generally reflects well (e.g. this and this. However, the article itself makes the claim in the LEAD that it's a newspaper of record and has sources to back it up. Is that sufficient for inclusion here? Best, Barkeep49 (talk) 01:51, 11 November 2020 (UTC)

Barkeep49, I think so. I don't think that the bar has been set very high for the amount of discussion necessary for inclusion here, as in my view this page is less the canonical catalogue of WP's perspective on a source (the way that RSP is used) and more like a scoreboard of all sources that have been discussed. _signed, Rosguill ^talk 02:16, 11 November 2020 (UTC)

I think you and I use this in slightly different ways, as I primarily use it through the Google Custom Search engine, while I think you use it more as a manual reference point when you encounter sources in articles. So I think I want to see a little more "oomph" behind inclusion but agree that there's enough in any case to include Globe and Mail. Best, Barkeep49 (talk) 02:19, 11 November 2020 (UTC)

OK to add external links?

Recently I started writing a citation highlighter user script that uses NPPSG. The script is working and highlighting Wikilinks to articles (e.g. The New York Times) based on their reliability, but NPPSG doesn't have very many domains (e.g. https://nytimes.com/ ). The citations at the bottom of pages (what we want to highlight) tend to use domains.

Would it be OK for me to start adding external links to NPPSG? Maybe something like:

• The New York Times [2][3], a newspaper of record in the U.S.
• The New York Times, a newspaper of record in the U.S. [4][5]

–Novem Linguae (talk) 13:19, 6 February 2021 (UTC)

Ping Rosguill. Definitely want to get your approval before making large changes. Thanks my friend. –Novem Linguae (talk) 05:03, 9 February 2021 (UTC)

Novem Linguae, I can't see a reason why not. Do add some clarifying text to the How to use this page section if you start adding links so that people don't confuse the roles of external links and citations to discussions. _signed, Rosguill ^talk 05:57, 9 February 2021 (UTC)

OK to import sources from WikiProjects?

For example, this WikiProject Film resources page has a large list of reliable sources and a small list of unreliable sources. Can I import stuff like this into NPPSG without citing RSN discussions? Thanks. –Novem Linguae (talk) 13:12, 8 February 2021 (UTC)

Yes, i think I did so for a few of the WikiProjects but never got around to doing them systematically. _signed, Rosguill ^talk 05:57, 9 February 2021 (UTC)

OK to add infobox websites?

Chemistry and medicine infoboxes contain some external links built into them. Examples: Ivermectin, Penicillin, Hydrogen iodide, Hydrogen peroxide. The fact that those websites are consistently used in very popular infoboxes seems to me like a strong tacit consensus as to their reliability. However, data is not a secondary source. Thoughts on including those websites as reliable at NPPSG? I'm leaning yes, since reliability is separate from primary/secondary/tertiary, but happy to hear other thoughts. Examples of websites that would be added as reliable: https://commonchemistry.org, https://chemspider.com/, https://guidetopharmacology.org/, https://drugs.com/ –Novem Linguae (talk) 06:36, 12 February 2021 (UTC)

Rosguill, hello my friend. If you have a minute, I wanted to get your opinion on the open issues on this page. The two sections at the bottom here, and the "Publishers" section a little higher up. Thank you for your time. –Novem Linguae (talk) 13:52, 14 February 2021 (UTC)

Novem Linguae, responded. Regarding this issue specifically, I think my perspective is the same as for the section below: reliable non-secondary sources are fine to include as long as we clearly identify their nature in the entry, but it would be best to find an actual discussion or documentation precedent for their reliability. _signed, Rosguill ^talk 17:16, 14 February 2021 (UTC)

OK to include legal primary sources?

Thoughts on adding legal primary sources such as https://justia.com to the list? It's reliable -- law FA's link to it extensively. But the fact that it's primary makes it a gray area for adding to this list. Being primary, it would not count toward passing GNG. –Novem Linguae (talk) 00:04, 14 February 2021 (UTC)

I think that as far as the primary status is concerned, we should just make sure that the entry is clear about that. My main concern is how we would cite their justification for inclusion, is there any documentation establishing the use of these sources (even an essay would be fine)? _signed, Rosguill ^talk 17:07, 14 February 2021 (UTC)

Sounds good. I'll go ahead and add important primary sources, but only with a reference. –Novem Linguae (talk) 17:39, 14 February 2021 (UTC)

Entries with only 1 vote, or 1 unclear vote, at RSN

I noticed a couple of these in the Europe section (Germany, Poland, Latvia, Russia). When these get added, they always seem to get added as "no consensus", regardless of what the RSN vote was. I wonder if it might be better to delete them, or to put them in the category that the voter suggested. For example, Neues Deutschland is in the "no consensus" category, but the only editor that commented on it at RSN compared it to Bild, which is unreliable. Thoughts? –Novem Linguae (talk) 15:28, 18 February 2021 (UTC)

Novem Linguae, my perspective is that one editor is not a consensus, but may still have useful insights, so we should generally list those in the no consensus category with a description of that editors' position. That having been said, I can think of some exceptions: a discussion that has only one editor give an opinion, but that opinion is backed up by solid evidence (e.g. CJR articles about the source, concrete examples of either major awards won or clearly deliberate misinformation), or a source that is so well known that no one seriously feels the need to address its reliability or lack thereof. In these cases, I think adding it to the Reliable or Unreliable category on the basis of one editors' comments is justifiable. _signed, Rosguill ^talk 16:07, 18 February 2021 (UTC)

Thanks for your perspective. I gave this some more thought. The thing I'm worried about is the "no consensus" headings on NPPSG list are actually a shorthand for "no consensus/yellow/marginally reliable". But by including these 1 vote sources as "no consensus", we are possibly giving them the wrong rating. Their real rating is "unknown", which is indicated for most other sources by not putting it on the list at all. The fix would be to remove the 1 vote sources, and have the standard be 2 mentions at RSN or similar to get added to NPPSG. For the moment, I created a workaround in my user script so that lines containing the words "insufficient discussion" or similar are not highlighted. –Novem Linguae (talk) 13:46, 22 February 2021 (UTC)

I agree with your assessment about how this category combines a few different types of discussions, but I would note that what all of these discussions have in common is that they 1) lack a clear community consensus on how to use the source (whether due to genuine disagreement or simply a lack of participation) 2) may nevertheless contain useful insights about the source in the discussion itself. Rather than split hairs abut what crosses the threshold, I think it's simpler to just include all possible entries in this category together and leave the entries as a resource for editors to do further digging, rather than gatekeeping and possibly removing useful information. _signed, Rosguill ^talk 17:05, 22 February 2021 (UTC)

Rosguill Hmm. Could also make a 4th category too, i.e. ; Insufficient discussion What are your thoughts on that, my friend? –Novem Linguae (talk) 03:01, 24 February 2021 (UTC)

Novem Linguae, that could work, but the line between insufficient discussion and no consensus is blurry, whereas I'm not really sure that an editor using this resource would end up treating these cases differently. For either category, general caution and reviewing the cited discussion are likely the appropriate courses of action. _signed, Rosguill ^talk 06:16, 24 February 2021 (UTC)

References

Wikipedia:News sources

I found the above page and associated effort the other day. I think these two sets of pages should be synced and/or redirected.

NPPSG is a good resource and looks more developed, but it 'lives' in the NPP space when it could be useful outside of NPP. OTOH I am not sure all the sources in NPPSG are in news sources (and anyway wouldn't want to split page histories per se).

Any opinions? --Izno (talk) 17:40, 18 February 2021 (UTC)

Izno, nice, looks like a good resource, thanks for sharing. One possible problem is that Wikipedia:News sources doesn't appear to have any WP:RSN reliability ratings. That's how NPPSG is sourced and organized, so it might be hard to add sources to NPPSG without those ratings. –Novem Linguae (talk) 13:50, 22 February 2021 (UTC)

@Novem Linguae: I'm kind of confused how this page is useful if it requires RSN reliability ratings and/or WP:RSP. Why not just look there? --Izno (talk) 17:07, 22 February 2021 (UTC)

Responding for Novem Linguae here, but the reason is that RSN's search function is lacking. This presents the results in a clearly readable format that makes the results of RSN discussions more accessible (and can also be scraped for use in scripts). The RSP entries are included for completeness, and are cited to RSP so as not to duplicate the effort of assessing discussions' consensuses. _signed, Rosguill ^talk 17:10, 22 February 2021 (UTC)

Izno, I'd be open to moving this guide out of "NPP space". While at first I was concerned that this guide could be misunderstood or misused if it was displayed as a general resource, with editors treating it as a bible of source ratings rather than a scorecard of how previous discussions have gone, enough editors outside of NPP have expressed enthusiasm about this as a tool to make me reconsider my original assumptions. _signed, Rosguill ^talk 17:10, 22 February 2021 (UTC)

I mean, if domains (music, video gaming) can do it, why shouldn't there be a general use reference for 'I can't read Swedish, is Aftonbladet generally reliable?' :^) --Izno (talk) 17:16, 22 February 2021 (UTC)

Custom search engine

Hey Barkeep49. Looking at the NPPSG talk page history, I see you created a Google custom search engine. Spreadsheet. Just wanted to let you know that I added external links to every entry in NPPSG, and also the list has grown from 200 entries to 1000 entries since you first made your search engine. If you are ever interested in updating your search engine, let me know and I can write a script to scrape the URL's. (In case you are wondering, I added all these URL's so that my CiteHighlighter user script could scrape it.) –Novem Linguae (talk) 15:56, 25 February 2021 (UTC)

@Novem Linguae I've noticed that you've been doing major source expansion here. I'm glad to see someone besides Ros and Interested in maintaining this. Currently the CSE stands at 309 entries roughly reflecting the state of the page in mid-December (I had forgotten about the spreadsheet and so it is not updated). If I'm being honest I'm a bit nervous about the 3 fold expansion you've done - you seem really interested in NPP (which is great) but you hold neither the NPR nor autopatrol PERMs and so I feel a bit of obligation to really check the validity of the additions. I'm not saying you've done anything wrong - my quick observations suggest you've done really valuable work. I am saying that you don't have our traditional measures that indicate you understand Wikipedia sourcing behind you at this time.

I have also displayed a bit of editorial judgement about what to include in the search engine. Rosguill has used a very low bar for including something from RSN which I think appropriate in a list form - if a reviewer is coming here to check a source they've already found such description might can be sufficient given that they'll also have made some evaluation of the source already. However the search engine is going to be used in other ways and so inclusion of marginal sources has a bigger impact because there's going to be less human judgement about how those results are then used. I have done similar things when including MUSICRS sources - some entries there don't include a link to a qualifying discussion so I have not included them in the CSE. And given the scope of this task and other pulls on my time I simply haven't had a chance to do that work yet. Truth be told I might not get to it for a bit longer - I suspect I'll need to devote several days of effort on this and so it might need to be done when I am on vacation from work. Thanks again for the work you've done here, Barkeep49 (talk) 16:13, 25 February 2021 (UTC)

Hey Barkeep49. No worries about the search engine. Just wanted to let you know that when you're ready, I'm happy to help write a script to scrape things. I could even write a script to only scrape lines that have a citation to RSN, and output it in a format that is ready to paste into your spreadsheet.

Sorry to make you nervous with my additions. Here's a quick rundown of my additions so that you can get a better idea of the quality control process I'm using:

• individual entries - I do RSN searches and I only add if I can find mentions by two separate people at RSN. I do not count the section starter's opinion. I do count passing mentions, and I do add together votes from different threads. I average the votes and link to the RSN discussions.
• MEDRS section - is mostly sourced from MEDRS guidelines and essays
• WikiProject lists - I am sourcing from WikiProject reliable source lists, when the WikiProjects have these. If the WikiProject only has a resource list rather than a list that says "reliable" and "unreliable", I skip it. There's a section on this talk page where I double checked.
• any mass imports or anything odd, I am double checking with Rosguill on this talk page. Lots of discussion like this toward the bottom.

As to my background, I am in NPP school with Onel5969, about halfway done. I joined AFC this week. Anyway let me know if any of the above is concerning to you and I'd be happy to adjust. I definitely want to stay in sync with you two. Thank you for your feedback. –Novem Linguae (talk) 16:51, 25 February 2021 (UTC)

Newspapers of record, preprint repositories

Hey Rosguill. Thoughts on me adding all the newspapers at the Wikipedia article Newspaper of record to this list as reliable? (Except for the North Korean one.) Also, thoughts on me adding all the List of academic preprint repositories as unreliable? I'd probably add these in their own sections, rather than merging them into other parts, seems easier and more organized. Looking forward to your feedback. –Novem Linguae (talk) 16:00, 25 February 2021 (UTC)

Novem Linguae, hm, that sounds reasonable, just make sure that the entries clearly identify that this is an RS calling the sources papers of record and not a Wikipedia discussion. I think that the newspapers of record should be integrated with the relevant countries of origin/coverage, however. The academic preprint repository suggestion sounds fine, although the entries should include a caveat that it is a preprint site and thus the same paper may have been published in a peer-reviewed publication as well. _signed, Rosguill ^talk 16:07, 25 February 2021 (UTC)

I would be against the preprint stuff being added. I think it is unlikely that a NPP will come to such a site and need help figuring out what is going on. The idea here is not to create a master list of all RS/non-RS, it's to assist new page patrollers as they look at new articles. And the number of new articles that are based on academic sourcing alone is a very small portion of the overall article base. Best, Barkeep49 (talk) 16:33, 25 February 2021 (UTC)

As a note, WP:UPSD already does this, and doesn't have what I consider a major downside: Labeling a source as reliable when it hasn't been checked in detail. NYT is generally reliable. That doesn't mean that it always is. It's in particular really not great for scientific and medical sourcing. This script would highlight it in green, misleading the reviewing into thinking the source is fine when it's not. Headbomb {t · c · p · b} 07:18, 1 March 2021 (UTC)

Headbomb, are you talking about my user script? It highlights regular RS and MEDRS differently (light green vs dark green). And highlighting generally reliable sources green is not a radical idea (e.g. WP:RSPSOURCES). But anyway, the user script has not been well received on the WikiProject talk pages I posted it at. And I am also getting a vibe that my edits to NPPSG have been too bold. I think a good solution going forward is I will make large additions to a private list instead of NPPSG. –Novem Linguae (talk) 08:08, 1 March 2021 (UTC)

Thailand

I was showing another editor this resource and they noted we don't have any entries for Thailand. I don't have tons of time to check it out myself right now but thought I'd post it here in case someone else is able to find some discussed sources for that country. Best, Barkeep49 (talk) 16:45, 21 April 2021 (UTC)

which section does JRank belong in?

It's clearly unreliable based on past RSN discussions (link) 137a (talk • edits) 11:55, 31 March 2023 (UTC)

IMO it seems to fall under Unclassifiable. _signed, Rosguill ^talk 17:01, 31 March 2023 (UTC)

How would you rate these sources based on past discussions?

• Canadaland ((archive 306)
• Student Crowd and IWMBuzz (archive 261)
• Nativeappropriations.com (mentioned at archive 261)
• CBN/Christian Broadcasting Network (archive 261#CBN - Christian Broadcasting Network)

137a (talk • edits) 15:46, 12 April 2023 (UTC)

I would say for the purposes of this list:

• Canadaland - no consensus
• Student Crowd - unreliable
• IWMBuzz - unreliable except for maybe ABOUTSELF-type information
• Nativeappropriations - no consensus
• Not seeing any relevant discussions at the CBN archive link

I tend to be a bit more epistemically conservative when assessing discussions for this list that have not been formally closed; a formal close could perhaps justify teasing out more of a consensus in the Canadaland discussion, but I don't think it would be appropriate to do so in a situation where the discussion's participants will not have an adequate opportunity to evaluate the close, and where the discussion was small enough that it would be conceivable that consensus could shift significantly with greater participation.

I'm not sure I'd bother listing that nativeappropriations discussion, as it's mostly just a fight between two editors over application of DUE and FRINGE, with no real insight into the source's reliability. The prior discussion had a bit more discussion of the source itself (mostly by me) but again the crux of the issue was a question of DUE across various subjects; I'd call that discussion a no consensus outcome as well (although I think I'd personally stand by my reasoning that it's reliable as an expert blog with use by others, if not necessarily DUE for every subject it writes on). _signed, Rosguill ^talk 16:14, 12 April 2023 (UTC)

Thanks. I got the CBN link wrong, here are the past discussions. Looks like "no consensus" to me. 137a (talk • edits) 17:04, 12 April 2023 (UTC)

Looking through some of the more recent discussions I'd agree. There's knee-jerk responses objecting to their use for anything controversial (and nobody seems to propose that we could use it in such contexts), but there's further genuine disagreement on the extent to which they are reliable for Evangelical topics, including in particular contemporary Christian music. _signed, Rosguill ^talk 18:09, 12 April 2023 (UTC)

Irish Mirror

Should the Daily Mirror entry extend to the Irish Mirror as well? I found no mention of it in past RSN discussions but I think that it's a sister paper of the former. 137a (talk • edits) 17:48, 31 May 2023 (UTC)

Rewrite to remove reliability classification for anything without RfC

Many of the sources are being judged by whomever added them to have been deemed "reliable" or "unreliable" based on, in the most egregious cases, a single RSN thread with a single editor's opinion. Consider the "unreliable" judgement for Pakistan's DND based on this RSN thread, and its nonsensical reasoning (editors can't be writers now??? Shall we ask CJR Editor-in-Chief Kyle Pope about this absolute ethical boundary?). RSN has deep problems as it is, but at the very least an RfC tends to get at least a few outside editors involved, keeps comments open for a while, and results in a definitive answer. Moreover, writeups and summaries on the WP:RSP are regularly patrolled -- they're usually not bad. Judgements made by editors on this page have none of these safeguards, yet from the comments here it would seem that editors will still be using this page as a guideline to definitively accept or reject sources.

I recommend that for any source on RSN that lacks an RfC, the listing be simply the source name (and any alternates and affiliates) with links to past RSN discussions and dates, without comment. Those sources with RfCs should be classified by listing the most recent RfC result with link and date, while also linking previous RSN and RfCs with results and dates. Thus it would look similar to WP:RSP, except tat these sources are not perennial. SamuelRiv (talk) 00:27, 15 August 2023 (UTC)

See Wikipedia:New_page_patrol_source_guide#How_to_use_and_improve_this_page. The intent of this page is not to provide a decisive rating to head off discussion, but rather to document where the most recent discussions have left off. _signed, Rosguill ^talk 19:34, 5 March 2024 (UTC)

in the most egregious cases, a single RSN thread with a single editor's opinion. I'd be in favor of removing the "single editor's opinion" ones, but will defer to Rosguill on this as the maintainer of this list. I use program code to look for keywords such as "one editor" to remove these from my user scripts. I would not be in favor of removing other cases, such as an RSN discussion with 2, 3, or 4 editors. –Novem Linguae (talk) 20:28, 5 March 2024 (UTC)

FWIW, my general practice is that "one editor" examples get a "no consensus" assignment. There's a few exceptions where the argument in question was a clear-cut case on the basis of something like WP:UGC or WP:NEWSORG grounds for papers of record. _signed, Rosguill ^talk 20:35, 5 March 2024 (UTC)