Wikipedia talk:Growth Team features: Difference between revisions

Short description task

I was thinking about a short description task. Some advantages and disadvantages:

Pro

• Easy to mark articles that need short descriptions
• Teaching a user to add short description is easier given their prevalence when using the search bar
• Can be modeled after WP:Shortdesc helper

Con

• Small peripheral contribution
• Policy for short description (content & when to add) probably varies between Wikipedias

What do you guys think? Sungodtemple (talk • contribs) 01:02, 6 March 2023 (UTC)

I like that thought! I think it'd be a great, easy task for entry-level newcomers, and probably easier to build compared to some other tasks. The one con I'd add is that it's a task where it'd be a little harder for newcomers to see the impact of their work, and thus might feel a little less fulfilling. The intro to the task would have to present very clearly, "this is where short descriptions show up," so that folks know what they're doing. Cheers, {{u|Sdkb}} ^talk 19:35, 6 March 2023 (UTC)

FTR, the Apps team works on this suggested task. Trizek (WMF) (talk) 15:48, 13 March 2023 (UTC)

Courtesy pinging @JTanner (WMF) and @ARamadan-WMF in that case. Cheers, {{u|Sdkb}} ^talk 15:55, 13 March 2023 (UTC)

I should have add added that English Wikipedia asked to have this feature being turned off. Trizek (WMF) (talk) 16:05, 13 March 2023 (UTC)

Oh, I took a deeper look at the page there, and it looks like there are a few different things here. Sungodtemple, as I understand it, is envisioning a tool that prompts and helps newcomers write SDs, not necessarily drawing from Wikidata. The previous feature that the community asked to turn off drew SDs automatically from Wikidata. And the apps team is currently looking at prompting users to add machine-generated SDs.

Overall, my stance is that this whole area is clouded by the original sin of forking Wikipedia SDs from Wikidata, which has created a ton of redundancy and will cause massive headaches when we eventually re-merge them. There are also complications from the fact that the community hasn't decided what exactly we want SDs to be. There are ongoing discussions about questions like how strict we should be about the 40-character limit, which biographies should have birth/death dates, etc. That lack of consensus means that, even if we could magically teach new editors anything, we haven't yet decided exactly what best practices we'd want to teach them.

It also makes machine-generated SD suggestions a particularly risky proposition at this point. For example, I'm curious, Amal, do the suggestions for biography pages currently include birth/death dates or not? That'll indicate perhaps which way the consensus is tipping, but we don't want to have newcomers adding these dates at mass scale until the best practices are firmed up. Also, if the guidance includes nuances/caveats, I worry that machine-generated SDs might miss them. For instance, the current guidance says that dates should be added only if the formatting criteria are met, which means we don't want them when adding them would push the SD over 40 characters. Do Descartes' suggestions reflect that consideration? If not, I think you might find it hard to get Wikipedians' consensus to use it.

Cheers, {{u|Sdkb}} ^talk 16:47, 13 March 2023 (UTC)

Hey @Sdkb, deeply appreciating these reflections. @ARamadan-WMFshared our project page on the improvements we are attempting to make to the apps Article Descriptions Suggested Edits task. As it stands right now, Android app users must have 3 edits before they can complete a suggested edit. We put this gate in because we agree with the sentiment that as it stands, article descriptions would be a challenging first task/edit. Machine Assisted Article Descriptions is an experiment in every sense of the word. I invite you to read the project page for the full details on our hypothesis, but we are aiming to understand if the algorithm improves the quality of article descriptions across languages, editor activity and tenure. English Wiki is one of the very unique cases of being divorced from Wikidata, which is why seeing how this would or would not be helpful on different language wikis is important. We will have experienced editors grade what is submitted and evaluate if people are editing what is suggested by the machine or submitting it right away. Please do let us know if you'd like to serve as one of the volunteer graders so we can capture views like yours in judging if this is a step in the right direction or not and what next steps should look like. JTanner (WMF) (talk) 23:52, 16 March 2023 (UTC)

Sure, I'll be happy to help out as one of the graders. Cheers, {{u|Sdkb}} ^talk 03:03, 17 March 2023 (UTC)

Hello,

My reply here will be only about apps, not Growth.

So, yes, the Android app team is working on improving the article descriptions suggested edits quality now.

You can check the mentioned above link "Machine Assisted Article Descriptions" for the whole clear information about the experiment background, objectives, and updates as well.

And for the workflow, here is the main Phabricator ticket, and from here, you can have a vision of how this will be presented to onboard the users on the feature usage.

Referring to the listed worries, I will make sure to get back to you with a reply after showing them all to my team.

With appreciation. ARamadan-WMF (talk) 17:55, 13 March 2023 (UTC)

Reminder about upcoming Growth team meeting

Hi all, the details of the upcoming Growth team meeting got a little buried, so I wanted to remind you all that it's happening tomorrow (Thursday, March 9th). Anyone interested in newcomers, Growth features, or the Editing team's Edit check project is welcome to come! I hope to see some new faces, along with others who have attended previous meetings! @Sage (Wiki Ed), @TheDJ, @Nick Moyes, @MB, @Clovermoss, @Sdkb, @SmokeyJoe, @Atsme, @theleekycauldron @Enterprisey @BusterD

• Date: Thursday, March 9
• Time: 17:00 UTC / 9:00AM Pacific Time (check timezone conversion here)
• Video conference link
• Notes

Agenda:

• Introductions (~5 mins)
• Kirsten: Quick Product & Technology annual planning intro + potential future Growth projects + Q&A (~10 minutes)
• Peter & Nicolas: Edit check demo + Q&A + feedback (~35 minutes)

Meeting Goals:

• Share an overview of where we are at in Product & Technology annual planning along with a quick overview some of the projects and priorities the Growth team is considering in the coming year.
• Equip volunteers with the context they need about Edit check to:
  • Identify things that could go wrong/might not work
  • Learn what questions and ideas this project brings up
  • Name implicit assumptions we need to hold ourselves accountable for validating
• Prepare this group to have future discussions about how Edit check might complement (or be extended) to address some of the Article creation for newcomer ideas this group is brainstorming.

KStoller-WMF (talk) 01:14, 9 March 2023 (UTC)

Meeting feedback

• Fun meeting! Very exciting to see tidbits and concepts which seem obvious but I hadn't considered. Edit check looks very powerful. BusterD (talk) 18:14, 9 March 2023 (UTC)

  Thanks, @BusterD, for the meeting feedback! I agree that these conversations have been enjoyable, and I know that Peter and I truly valued the feedback we received in this last meeting! Hopefully I'll have more details to share soon on how Growth team plans will be shaped by annual planning priorities, but no matter which project gets prioritized I definitely value continuing to have meetings (as long as it works for you all). I'll reach out to you and other previous attendees in a few weeks to suggest a new meeting date and time. I'm thinking a reoccurring meeting every other month might be a good cadence? KStoller-WMF (talk) 23:21, 17 March 2023 (UTC)

In Starship Troopers (film), there's a buzz line used over and over: "Do You Want to Know More?" (always accompanied by an onscreen "yes" button); I was proposing we start asking the question: "How Do You Know?" (accompanied by a positive response button). I've been thinking about younger subscribers for a while. (I'm trained as an elementary school teacher so guess why I jump to that.) "How do you know?" almost sounds like the fourth-grade version of the "why?" question a three-year-old will repeat over and over. This is a good thing. How do you know is a very healthy question to ask. At any age, on any subject. BusterD (talk) 18:25, 9 March 2023 (UTC)

I like this. Certainly inquisitive "How Do You Know?" language is far more positive and proactive than the current revert / reactive approach that is more common now. I'm not sure what copy Edit check will settle on, but I know that Peter is hoping to make the feature community configurable in the future, and perhaps that might even apply to the language used in the notice. KStoller-WMF (talk) 23:28, 17 March 2023 (UTC)

• I wanted to mention that I wrote a comment here about some of my thoughts, mostly as a response to comments PPelberg (WMF) made. Clovermoss🍀 (talk) 18:28, 9 March 2023 (UTC)

Meeting Summary

Thank you, Everyone, for joining our call on March 9th, 2023. Here is the link to the recording, and below is a brief summary:

• During the first half of the call, Kirsten Stoller, Senior Product Manager for the Growth team, presented a high-level overview of the Product & Technology annual planning process and potential projects for 2023/2024.
  • We discussed the possible risks of a large number of extensions not having active maintainers and how some of the proposed new features might not actually lead to newcomers being ready to create new articles.
  • For follow-up questions, you can:
    • add comments or questions to this document
    • reach out to Kirsten on Wiki: KStoller-WMF
    • send an email: kstoller@wikimedia.org
• For the second half of the call, Peter Pelberg, Senior Product Manager, and Nicolas Ayoub, Senior UX Designer for the Editing Team demoed the latest progress for the Edit Check project
  • In general, there were positive reactions and excitement about the project.
  • We discussed some of the potential challenges for mobile editors and use cases with bad actors.
  • Next steps: Peter and Nico will revisit the comments and respond to everything that hasn’t been addressed during the call.
• In the future, we are planning to start hosting monthly Growth office hours to have similar discussions with the wider group.

--MShilova (WMF) (talk) 21:44, 9 March 2023 (UTC)

When we last met we discussed several of the possible projects that the Growth team might work on next fiscal year, but we also mentioned that the Annual Plan wasn't finalized and that it would guide Growth team work. So I wanted to let you all know that the Product and Technology department's Annual Plan Objectives and Key Results is available for review and discussion.

From the Growth team perspective, an important detail to note is there isn't a key result specifically about new editors or new editor retention. So the Growth team may work on project related to the "Contributor experience" objective or a "Reading and media experience" objective. It's also possible that some of our team capacity will be devoted to the "Future Audiences" focus.

Throughout April and May, you are invited to leave comments or questions on the talk page, next to the subheading associated to each draft Key Result. Some of the individual draft Key Results will also have associated "focus group" meetings, where the relevant Wikimedia Foundation staff team would like to talk with a small group of people who care about that topic in particular, to discuss the topic in detail. I'll start a new discussion and ping previous Growth team meeting attendees once I know more details about focus groups that relate to the Growth team!

Also, the draft annual plan for the whole Wikimedia Foundation is available here. There will be several synchronous and asynchronous ways to provide feedback to that plan.

Thanks! -- KStoller-WMF (talk) 21:49, 20 April 2023 (UTC)

Growth team newsletter #25

Welcome to the twenty-fifth newsletter from the Growth team! Help with translations

Celebrations

Leveling up release

• We released Leveling up features to our pilot wikis on March 22 for an initial A/B test.
• In this test, we use post-edit dialogs (pop-ups shown after publishing an edit) and notifications to encourage new editors to try new types of newcomer-friendly suggested edits.
• We are closely monitoring the short term impact of this feature as well as the longer term effect on newcomer productivity and retention. If the experiment shows positive results, we will release this feature to more wikis.

5,000+ images added via the newcomer task in February

• In February 2023, 5,035 images were added via the newcomer “add an image” feature (on all wikis where available); 155 were reverted.
• Since the feature “add an image” was launched: 36,803 images have been added; 2,957 images were reverted.

Recent changes

• Add a link
  • Community Ambassadors completed an initial evaluation that confirmed that prioritizing underlinked articles resulted in better article suggestions. We then evaluated the change on Growth pilot wikis, and results suggest that more newcomers are successfully completing the task and experiencing fewer reverts. We have now deployed the new prioritization model to all wikis with "add a link" enabled. [1][2]
  • We continue the deployment of "add a link" to more wikis. These changes are regularly announced in Tech News. To know if newcomers at your wiki have access to this feature, please visit your Homepage.
• The Impact module was deployed on our pilot wikis, where we conducted an A/B test. We published initial findings, and a data scientist is now completing experiment analysis. [3]
• Donor Thank you page experiment – Donors land on a “thank you” page after donation, and that landing page now includes a call to action to try editing: Example Thank you page in French. This promising feature is tested at several Wikipedias (French Wikipedia, Italian Wikipedia, Japanese Wikipedia, Dutch Wikipedia, Swedish Wikipedia).
• Growth features are now the default experience on both test.wikipedia.org and test2.wikipedia.org. You can test our features there.

Upcoming work

• Add an image – We plan to offer section-level image suggestions as a structured task for newcomers.
  • We are starting design work. [4]
  • Growth team Ambassadors helped with an initial evaluation of these section-level image suggestions. [5]
  • We will start a community consultation in which we share initial ideas and designs soon. [6]
• IP Editing: Privacy Enhancement and Abuse Mitigation – We will support this project for all Growth Team maintained products and extensions that may be affected by IP Masking. [7]

Growth team's newsletter prepared by the Growth team and posted by bot • Give feedback • Subscribe or unsubscribe.

13:10, 1 April 2023 (UTC)

Meeting with the Growth team and Moderator Tools team: May 4th

@Sage (Wiki Ed), @TheDJ, @Nick Moyes, @MB, @Clovermoss, @Sdkb, @SmokeyJoe, @Atsme, @theleekycauldron @Enterprisey @BusterD @Novem Linguae

Since our last meeting, the Wikimedia Foundation's draft Annual Plan has been published and the Product and Technology's draft OKRs have been published. Both are still subject to change, as they are drafts and will be shaped further by the community discussions happening on those talk pages.

The Growth team and the Moderator Tools team hope to meet with community members to talk through some ideas related to Contributor experience and improving moderation workflows. To achieve this, we would like to gather insights and opinions from moderators regarding the challenges they face. We're particularly interested in understanding more about how you determine where your attention is needed, and how you prioritize different tasks. When we say 'moderators' we're referring to editors with advanced rights, such as patrollers, administrators, and functionaries.

If you are interested, you can see meeting details and sign up here.

I hope to see some of you there! Thanks, - KStoller-WMF (talk) 20:07, 27 April 2023 (UTC)

Will Star Wars be involved? ― Blaze Wolf^Talk_{Blaze Wolf#6545} 20:14, 27 April 2023 (UTC)

Ha! Extra credit to anyone who shows up in a Star Wars costume?  ;)

Oh, I should have also mentioned there are several other calls and places to leave feedback, including on the talkpage of the OKR draft itself. - KStoller-WMF (talk) 22:29, 27 April 2023 (UTC)

Article creation hypothesis

Hello

Following the previous topic, just above, the Growth team is exploring a project idea that aims to improve the experience of new editors by providing them with better guidance and structure in the article creation process. The hope being that by providing new editors with more structure around article creation, it will lead to newcomers creating fewer low-quality articles that create work for patrollers who check recent edits and mentors who review newcomers’ drafts.

In 2022, about 28% of newly registered users who completed the Welcome Survey indicated that they opened an account specifically to create a new article (all stats). These newcomers don't yet understand core Wikipedia principles and guidelines around notability, verifiability, conflict of interest, neutral point of view, etc. These newcomers need additional guidance or they end up frustrated and disappointed when their articles get deleted. Because they aren't receiving the proactive guidance they need, they end up creating additional work for content moderators (patrollers, admins, watchlisters…) who need to provide reactive guidance which is rarely well-received or well-understood.

While the specifics of the project, and the Growth team’s annual planning priorities, are still under consideration, we anticipate exploring ideas related to  Article creation improvements for new editors.  One possibility is a community configurable "Article wizard" or helper, which could also fulfill the 2023 Community Wishlist survey Reference requirement for new article creation proposal (ranked #26 out of 182 proposals).

We're committed to shaping the overall plan based on community feedback and needs, while adhering to the following requirements:

• The feature will be Community configurable, enabling each community to customize it to meet their unique needs.
• The feature will provide guidance and guardrails to help newcomers create higher-quality articles and improve their overall experience.
• The feature will be designed to reduce the downstream workload for content moderators.

So, we would love to hear from you:

1. Do you think this project will help new page patrollers on English Wikipedia?  
2. Do you have any suggestions for improving this idea?
3. Is there anything about this idea that you find concerning, or you want to ensure we avoid?

Or do you want the Growth team to consider a totally different idea?  Keep in mind that the Moderator Tools team and two other teams are also working the shared  “improve the experience of editors with extended rights” key result, so there will be other teams approaching this from a less new-editor centric perspective.

Thoughts? :) Trizek (WMF) (talk) 18:25, 15 May 2023 (UTC)

@Trizek (WMF) ultimately I would say this comes down to two questions:

1. Will the "harm reduction" of improving those editors who are going to make an article first thing regardless outweigh the risk of the tool nudging editors who wouldn't make an article (or could be persuaded to delay) into trying article creation out early on?
2. What is the opportunity cost of creating something like this?

In en-wiki (and likely other large wiki) terms, I couldn't begin to make a guess as to how the trade-off will play out, but I could well see something like this really helping out on wikis that are still in a prioritise new article stage, likely with lower thresholds to be met by new articles. Something that ran new editors through the basic questions (including catching unintentional paid-editor violations) and guidance could well be worthwhile. I actually suspect it would probably help AfC more than NPP.

In terms of suggestions, things like getting them to go find sources first, and efforts to try to trim out as many unsuitable sources as possible (whether through guidance, or machine-based through reference to WP:RSP and the blacklists, or a mix) would say a lot of aggro.

In regards to the opportunity cost - if Growth didn't do something like this, would would make it in that otherwise might not? Nosebagbear (talk) 14:32, 16 May 2023 (UTC)

+1 on all this. To expand on the risk of the tool nudging editors who wouldn't make an article (or could be persuaded to delay) into trying article creation out early on, creating a new article is inherently an advanced task, since it requires mastering all the different elements that go into an article, whereas other tasks often require only mastering one. That'll stay the case even if the Article Wizard is optimized to perfection.

So if our goal is to ultimately recruit more editors who go on to become senior contributors, there's a risk associated with nudging them to create an article too early, leading them to get overwhelmed and quit. Among the 28% who join Wikipedia intent on creating an article, I'm not sure how much potential there is for recruitment anyways — many of these are SPAs who care only about promoting the entity they're trying to create an article about, not Wikipedia's broader mission of documenting knowledge. Even if we're not recruiting future admins, there's still some potential value: content created by such editors can be decent if they're sufficiently corralled into abiding by neutrality rules, and if the project makes it easier for patrollers to deal with these editors then it could reduce our workload. But as you're weighing whether to pursue this or not and looking at that 28% number, I'd just urge you to keep in mind that not all newcomers have equal potential to develop into experienced editors. The group intent on article creation may be a large pool, but the quality candidates in it are probably a lot more spaced out than in some other smaller pools.

Best, {{u|Sdkb}} ^talk 17:34, 16 May 2023 (UTC)

If someone's single purpose is to create an article about a notable topic and a wizard lets them do that better than they could on their own, I think ultimately Wikipedia is a little better off than they were before even if they disappear afterwards. If someone's single purpose is to create an article about a non-notable topic and this lets them do that better than they could on their own, that becomes a problem because it will mean more volunteer time is having to be spent reaching the conclusion "this is not an appropriate topic to cover on enwiki". So how much of that 28% fall in the "notable topic, lack skills" category and how much falls in the "non-notable topic, lacks skills" category really matters in figuring out the net positive/net negative impact of this. Best, Barkeep49 (talk) 18:35, 16 May 2023 (UTC)

@Nosebagbear and @Sdkb, Thank you for providing your insight! I believe it's important to clarify that this project does not aim to funnel more newcomers into article creation (at least not on larger wikis). Growth tools encourage new editors to start with smaller edits, and we do not intend to change that. We are simply thinking about ways to provide better guidance for newcomers who progress to article creation early on and really need more help, especially around providing reliable references.

Regarding the opportunity cost of undertaking this project, we are currently evaluating the impact, effort, and risks associated with various project ideas. Another project idea in discussion is to make the newcomer homepage more modular and "grow" with editors as they become more involved and experienced. In fact, @Johannnes89 recently suggested a similar concept to the Growth team. If communities express interest, we can certainly consider this project instead.

There is also the possibility that the Growth team will work on a project related to a different annual plan key result: potentially a project more centered around readers.

In light of this information, do you have concerns with the Growth team focusing on providing more structure for editors creating their first article or article draft? Or do you have thoughts on other priorities the Growth team should focus on that is still in alignment with the Contributor experience, key result 1.2? KStoller-WMF (talk) 18:22, 19 May 2023 (UTC)

I really like the idea of adding more reader-focused features to the homepage. Wikipedia's massive readership is such a huge asset, and if we can convince even a tiny fraction of it to create an account, that'll be a huge number of users at the (very) start of the path from reader to editor. {{u|Sdkb}} ^talk 19:34, 19 May 2023 (UTC)

I know NPP have had some discussions with the moderator tools team about this and think it can be useful no matter which team does it. One important note from a Growth Team perspective: newly registered editors cannot create mainspace articles, they have to do so in draftspace. So there is going to be some level of barrier/impediment for the 28% of new editors even if this project is successful. This is copied from WT:NPR where I originally saw this announcement to here so that any discussion here among non-WMF people can consider it. Best, Barkeep49 (talk) 14:45, 16 May 2023 (UTC)

Thank you for your replies.

@Nosebagbear, the idea behind this project is indeed to help wikis getting more quality contents. Biggest wikis have more rules and expectations, that could be conveyed in the article creation process, while smaller wikis would be happy to get more articles being written with an increased quality. I'll post the same message to AfC, good idea.

Regarding the cost opportunity, well, at the moment, everything is possible. The goal of this discussion, which happens as well at ar, bn, cs, es and fr, is to get more people feeding our discussions.

@Barkeep49, NPP discussions are one of the reasons why we consider this project for the upcoming fiscal year. Regarding the article creation process, going through draft submission can be a step in article creation. Consider this project as a way to reduce the drafts backlog, as the quality of submitted articles would improve.

Trizek (WMF) (talk) 15:40, 16 May 2023 (UTC)

@Trizek (WMF): despite being the editor who consolidated the three parallel copies of this discussion, I'm not actually sure where to reply, and am low-key hoping someone with like a keyboard will refactor the combined conversation to make it easier to follow.

This isn't exactly an answer to your questions, but I believe the most important thing to hammer home in newcomer education regarding new article creation is the priority of sourcing, and here I'm using priority not in the sense of importance, but in its chronological meaning of firstness. In my opinion probably the best essay for newcomers to read before they try to create an article is WP:BACKWARD.

If I were working on improving / recreating the article creation wizard, the first interface screen would ask for the editor's best source, with a second field for them to estimate length of topic coverage in units of paragraphs or pages, nothing more granular. Then I'd have two sidey-side columns of tickboxes related to reliability and independence (not a forum post, not an interview, not a press release, not user-generated content, or whatever people think best encapsulates the most common stumbling blocks in these two categories). The editor would have to tick all the boxes to proceed. While this could possibly then produce a formatted citation with important fields like title, author, and date left invitingly empty, nothing would have to record or attempt to verify the input: the important bit is having the newcomer think through why their source is good enough, not to restrict their contributions.

After two of these "best source" interface screens, then I'd display an input field for what the editor wants to name their new article. If we can convey the idea that editors shouldn't even consider what their article is going to be titled without first finding sources for it, I think it would teach a lot without requiring much reading. And while we don't want to increase the barrier of entry, embracing this sort of "fail faster" approach should allow newcomers to work on crucial skills before spending time and effort writing a bunch of prose that gets frustratingly deleted due to unsuitability.

Having said all that, I don't work NPP or AFC, so my opinion should be understood as that of a partially informed outsider. Folly Mox (talk) 18:22, 20 May 2023 (UTC)

This discussion is currently taking place in three places simultaneously, with mostly non-overlapping participants. Since the present instance of the discussion is the one linked from the mediawiki page, I'm announcing my intent to copy-paste the other two instances here using the appropriate templates, to defragment the conversation. Folly Mox (talk) 15:46, 20 May 2023 (UTC)  Done 15:59, 20 May 2023 (UTC)

Moved from Wikipedia talk:WikiProject Articles for creation § Article creation hypothesis

 – Defragmenting conversation.

Just noting the parallel discussion about updating the existing Article Wizard. Also, it sounds like this project will help Articles for Creation, not to get too pedantic about who you're informing of your goals ;-)

But yes, overall this sounds like it would be a boon to this project, WP:NPR, and the Encyclopedia as a whole. Primefac (talk) 16:30, 16 May 2023 (UTC)

@Primefac, thank you! If you see it as a bonus, can I ask you which traps this project could fell in? Just to have a counter point of view, that will certainly feed the discussion! :) Trizek (WMF) (talk) 14:42, 17 May 2023 (UTC)

Bonadea hits on a few of them below; we want to make this easier for new users but not at the expense of making it harder for the reviewers. I honestly couldn't say what sort of pitfalls there will be, because we've been trying to improve the AFC process since before I joined and we are still coming up with ways to improve our processes (I think there are at least three discussions ongoing now about the matter). Some problems can't be solved, but as long as whatever is done can be easily undone if necessary, there is no real long-term harm. Primefac (talk) 08:24, 18 May 2023 (UTC)

It is depressing to see those stats. Not surprising, because so many drafts are created by new users who have no other edits, but profoundly discouraging.

I guess the first question is what makes this proposed article wizard better than the current one in en.wiki?

Looking at the community wishlist discussion, mention is made of a video explaining how to add references; I'm sure that is a nifty idea, but please make sure that no information is only available in video form. It has to be available as written text as well.

That any new article at en.wiki must have references is true, but this is not going to "reduce the downstream workload" (sorry, that kind of marketingspeak makes my skin crawl a little) for "content moderators" (unfortunate choice of words – we try pretty hard to make it clear to new editors that English Wikipedia has no "moderators", we are all simply editors). AfC reviewers and NPR patrollers will have to check the references in drafts / new articles just like we do now, and what takes time (and "increases the downstream workload") is not making sure there is a footnote in the text (that takes no time at all!) but explaining to the new editor why certain sources can't be used, why it's unacceptable to dump references at random points in the article or to place them all at the end of the article, why ten crappy sources are not better than no source, why ten good sources are not better than one good source in order to verify a simple statement of fact, etc, etc, etc. That's not going to change with a new article creation tool which requires one reference in order to publish an article.

The situation is going to look very different in different Wikipedia versions, but I do not have a sense that new articles without any source at all is a particularly big problem at en.wiki. Most of them do have at least one source, so I'm not sure if this would solve anything for this particular wp version. But I might be very wrong about that, of course. --bonadea contributions talk 18:55, 16 May 2023 (UTC)

@Bonadea, thank you.

We always keep thing in text format, as it is the less data consuming form to transmit information. Not everyone has fiber-optic high-speed internet connexion with a state-of-the art computer to edit the wikis.

Regarding the terms used, which one would you suggest to describe all users who loot at new changes, check submitted drafts or pending changes, etc? :) I almost went with "curators" but I've been told previously that it wasn't specific enough.

The "marketingspeak" is using vague terms for a reason: we want to hear your ideas, based on your experience, not to hear comments on ideas we have (not at this step). Let me sketch a couple of possible ideas (but keep in mind that nothing has been set so far, they are my ideas as an article reviewer at French Wikipedia): as you illustrated it well, reviewers have a lot of work to check on references. What if we help with a tool that pre-checks references when article creators add them to their draft? Or a tool that ask them if the footnote covers what they wrote in the last sentence? Providing content creation advice would reduce the number of non-compilent references. Again, I really don't like to share ideas, as they could influence the discussion, so please keep my disclaimers in mind.

Thank you for keeping other Wikipedia in mind. We started this conversation with 6 communities to have different perspectives.

Thank you again for your detailed reply. Trizek (WMF) (talk) 14:41, 17 May 2023 (UTC)

One step I think we could take to help the workload is make it much clearer to new editors what things are completely inappropriate for Wikipedia (junk like Draft:Thomas Mills for instance). New editors trying to make BLPs is one of the biggest potential areas we can have serious problems as far as article creations. I think what you are proposing is fundamentally a good idea that could help out both NPP and AfC. Trainsandotherthings (talk) 01:19, 17 May 2023 (UTC)

@Trainsandotherthings, thank you! How can we make clear which topics are appropriate (or accepted)? Trizek (WMF) (talk) 14:45, 17 May 2023 (UTC)

We should be clear that certain things aren't appropriate at all (opinion pieces, autobiographies, etc) and in my opinion caution against BLPs for brand new editors in general. If we could get every new editor to read WP:NOT, we would save a lot of time. Trainsandotherthings (talk) 20:15, 17 May 2023 (UTC)

Notified WT:NPP/R. Skarmory (talk • contribs) 04:09, 20 May 2023 (UTC)

They already know, for what it's worth. Primefac (talk) 06:21, 20 May 2023 (UTC)

I didn't know that page existed, honestly. Not sure if I just missed it or if it's somewhat hidden, but I would've checked it if I knew it existed. Skarmory (talk • contribs) 10:20, 20 May 2023 (UTC)

Oh, not surprising, NPR has more pages than we used to! I just wanted to make sure everyone knows where all of the discussions are. Primefac (talk) 10:26, 20 May 2023 (UTC)

Moved from Wikipedia talk:New pages patrol/Coordination § Article creation hypothesis

 – Defragmenting conversation

IMO the fix needs to come from En Wikipedia. Sincerely, North8000 (talk) 18:32, 15 May 2023 (UTC)

I know NPP have had some discussions with the moderator tools team about this and think it can be useful no matter which team does it. One important note from a Growth Team perspective: newly registered editors cannot create mainspace articles, they have to do so in draftspace. So there is going to be some level of barrier/impediement for the 28% of new editors even if this project is successful. Best, Barkeep49 (talk) 18:37, 15 May 2023 (UTC)

@North8000, how can en.wp fix it, and how can the Growth team help there?

@Barkeep49, as dais elsewhere, submitting a draft can be a step in the process. Even if it is mandatory at en.wp, it also exists at other wikis, under a "strongly encouraged" advice.

Trizek (WMF) (talk) 15:42, 16 May 2023 (UTC)

Much can be accomplished by mere education about wikipeda and guidance. The underlying structural problem preventing moving forward on this is that functionally we only have two kinds of pages:

• "Rule" pages (policies and guidelines)
• Thousands of random essays and essay-like pages

We need two new vetted categories:

1. Vetted information and teaching pages Should be a short highly vetted list. Maybe 20 within a year and eventually never over 100
2. Definitive statements about Wikipedia. Ultra-consensused. "Constitution" pages; too vague to be invoked directly but they guide policies and guidelines. To start wp:5P would be the one and only of these. Eventual max of only 5 or 10

The foundation needed here is #1. We really don't give newbies guidance because the only guidance that we give is "Here's thousands of good and bad, useful and non-useful essays.....go randomly wander amongst those and try to learn". So one document under #1 would be the highly vetted official landing and navigation page for newbies which would only navigate to a short list of other #1, #2 documents and policies and guidelines. For "making a new article" it would route to another official #1 page which starts with the "should an article exist?" question and points them to read WP:Not & WP:Notability. It instructs them that step 1 of building (or deciding not to build) a new article is to find 1 or 2 GNG references and start the article with those. If it has those, chances are that the article should exist, regardless of how badly done it is. Voila, we have solved a whole bunch of new article, NPP, AFD problems and eliminate tons of drama and wasted work.

The other easy big fix is to convert AFC into just handling "should this article exist?" questions instead of the very tough article perfection gauntlet that it currently is. Then (and not until then) we need to nudge all newer editors to go through AFC. North8000 (talk) 16:16, 16 May 2023 (UTC)

Thank you for your detailed reply, @North8000.

I went through the AFC, and it seems to me that the most important elements to have in an article are detailed there. It is quite short. When you mention the numerous pages users are encouraged to read, I guess it is though the multiple links provided?

How to you imagine the "should an article exist?" process?

How a user would know if a reference proves notability?

Thank you again, Trizek (WMF) (talk) 12:59, 17 May 2023 (UTC)

• I assume you're aware that we already have an article wizard? Though it could certainly be improved. – Joe (talk) 13:11, 17 May 2023 (UTC)

  We are @Joe Roe. :) Plus, this discussion is not only happening at English Wikipedia, but at 5 other wikis that have different AFC processes. Trizek (WMF) (talk) 14:17, 17 May 2023 (UTC)

  So would this replace the existing wizard, enhance it, supplement it...? It's hard to answer your questions without knowing how your idea will fit into existing processes here on enwiki. – Joe (talk) 14:58, 17 May 2023 (UTC)

  @Joe Roe, I know it is difficult, but we want to hear you regarding article creation in general. We don't want to say "we will improve this" or "we will built something different", as it would left aside some ideas that would not be considered if we narrow down the context or the possibilities. What would help new pages patrolled at your wiki? What is overall missing in new pages? Trizek (WMF) (talk) 13:01, 22 May 2023 (UTC)

Here's the AFC problem and I'll use and extreme case for clarity. Let's say that an AFC article has 1-2 references that look like GNG references and complies with wp:not, but needs a lot of work which means it has a lot of big problems. IMO that article should be passed out of AFC into article space for further development but in reality it won't because a typical reviewer will not want to put their stamp of approval on such an article. So it's overly difficult to get an article through AFC. The solution is to make it clear that the only thing AFC reviewers are responsible for is "should this article exist" type questions. And I was arguing for a new pattern and new guidance saying in essence "Step one of creating a new article is to find 1-2 GNG type references". North8000 (talk) 17:02, 17 May 2023 (UTC)

Regarding "How a user would know if a reference proves notability?" My answer would be to just see if it has 1-2 GNG type sources that look likely for wp:notability. Say "it's OK to pass the edge cases". Edge cases can get handled later.North8000 (talk) 17:07, 17 May 2023 (UTC)

My sense is that AI is still a far way off from being able to automatically assess the "GNG-contribution" of a given source, let alone a set of sources collectively. The only idea that comes to mind tools to assist new editors in this process would be to implement a checklist that asks new editors to go through their cited sources and identify whether each one, individually, 1) contains significant coverage of the topic 2) is independent of the subject 3) is a secondary source, before letting them submit the page for review. If the editor's responses can be transcribed to the talk page or as an AFC comment, it could further aid reviewers in assessing articles that contain mountains of sources. _signed, Rosguill ^talk 17:19, 17 May 2023 (UTC)

@North8000, "should this article exist" will be a qualified yes for someone who wants to create it. What is the best way to explain that Wikipedia is not accepting all topics?

The following question is also for you, and @Rosguill (and everyone else, of course): do you think users who just create an account understand what it "a good source"? I have the feeling that the 3 conditions Rosguill lists are concepts that aren't familiar with most people.

Trizek (WMF) (talk) 13:09, 22 May 2023 (UTC)

1) I went ahead and informed Wikipedia talk:Article wizard.

2) May I also suggest in the future that the discussion be held on one page, and then additional pages be notified with the {{subst:Please see}} template? It seems the discussion is being scattered across a couple different pages.

3) The current article wizard ends with the new article being placed into draftspace, with a "submit" template included at the top of the draft, which I think is good. That part of the workflow should be kept.

4) The top 2 things about writing an article that give a new editor a bad experience, in my opinion, are A) writing an article on a non-notable topic that then gets declined or deleted since it is non-notable, and B) writing an article that is of borderline notability that then sits in the draftspace queue for 3 months because it is not an easy accept and not an easy decline (easy accepts and easy declines are processed quickly). This is often combined with WP:REFBOMBing. The delay in accepting/declining comes from the fact that the reviewer needs to click open all these 10, 20, 30 sources and evaluate each for GNG, which is a bit laborious and requires skill.

Both of these problems would be solved by getting newer editors to include multiple top quality, WP:GNG passing sources in their submissions, preferably at the top of the list of citations, and not drowned out by excessive other citations. But GNG is hard to teach to new editors. In my opinion, GNG is written quite vaguely, and attempts to add more detail to the guideline are declined. In my opinion, simply reading notability guideline pages will not teach a newer editor enough about notability for them to be able to determine if a topic or their article is notable.

I think there may be opportunities to insert a step into the article wizard that explains enough of WP:GNG to help out draft writers. Something like "Not all topics qualify for a Wikipedia article. The articles most likely to qualify will have citations to multiple high quality sources such as newspapers and books going into multiple paragraphs of detail about the topic. Before you spend a lot of time writing an article, do you have at least X such sources? Yes/No"

In fact I may propose this be added to the current article wizard.

5) Changing gears, I am not sure what a major overhaul of the article wizard would look like. Do we have a list of things we don't like about the current article wizard? Things that need fixing? Things we'd do differently? I think drilling down into these kinds of details, and then using them to create a detailed proposal, will be important for moving forward with any kind of article wizard overhaul. Hope this helps. –Novem Linguae (talk) 22:08, 18 May 2023 (UTC)

Thank you @Novem Linguae, it elps.

A quick note regarding using ==Discussion at ==

 You are invited to join the discussion at . Trizek (WMF) (talk) 13:18, 22 May 2023 (UTC): we observe that users are less likely to participate of we ask them to go to a different page than the one they read. :)

I took good note of your suggestion of an initial warning. Allow me to challenge it: all users who really want to create an article will say yes even if their article topic is not notable, and less confident users will click no even if their article topic is notable. How can we avoid this effect?

We can consider this discussion as a way to re-thing the Article wizard, even if we have no plans regarding improving or replacing it right now. If a list of WP:WIZARD improvements exist, I'm curious to read it as well. Trizek (WMF) (talk) 13:18, 22 May 2023 (UTC)

Agree with these posts. I think that even just giving some guidance that finding GNG sources is step 1 of making an article / deciding whether or not to make it and explaining / checklist regarding what a GNG source is. We don't need perfection here or to be overly stringent to worry about edge cases. BTW my comments were focused on making AFC more focused on notability and to ease up in the other areas. The it becomes more viable to nudge newbies to go through AFC. North8000 (talk) 11:54, 19 May 2023 (UTC)

Can we merge this discussion to WT:AFC#Article creation hypothesis to centralize it to one place? Having separate discussions on separate pages about the same thing seems like a bad idea. (I say to take it over there because that talk page is more active, but I don't really mind which one it ends up on; we could merge the discussion from there over here as well, I don't care.) Skarmory (talk • contribs) 12:26, 20 May 2023 (UTC)

Up to Trizek (WMF). I'd be fine with a merge. Maybe use the templates {{Moved from}} and {{Moved to}}, then cut and paste the relevant text. –Novem Linguae (talk) 13:28, 20 May 2023 (UTC)

Truthfully I think the one page it should be on is the Growth team's project page on enwiki because that's where updates will presumably happen. Best, Barkeep49 (talk) 15:41, 20 May 2023 (UTC)