Research:New editor support strategies/AFC p4

Editor has worked at AFC for several years, and been on-wiki for 4-5 years total (on and off).

Session video (YouTube)


Tell me about your AFC work?

I do a little bit of everything. The reason I probably showed up in your query is because I moved a bunch of stuff from the Wikipedia talk namespace into the Drafts namespace. But I also review AFC drafts on occasion. It's work that needs to be done, and work that's often backlogged.

How frequently would you say you review AFC drafts? How many a month?

On and off, I'm going to have an average/mean of somewhere between zero and one a month, maybe up to two. But it's going to vary between zero for most months and sometimes up to 1 a day when I'm really doing it.

That's the problem with the mean... not always the most representative statistic.

No it's not.

When did you start working in AFC?

I did it a couple of years ago and then I took a fairly long break and then restarted doing it just a couple of weeks ago. My contribution log would know a lot better than I would.

Okay, and did you stop... when you say you took a break, did you just stop working on AFC, or on Wiki in general?

Wikipedia in general. I was in college at that point, and I would correct an error if I randomly came across it, usually as an IP, but I didn't really actively edit in the project side.

And when did you start work on Wiki, the first time? How long ago?

Let me actually pull up my first IP edit. Do you have access to my screen?



So... I've got some contributions it looks like going back to 2011. I edited as an IP occasionally before that, so, would be probably fair to say that I've edited since mid-2011 on and off.

How, when did you become involved in AFC?

In AFC specifically? I think it would be about 2013. But...

Approximate is fine. So tell me a little bit about—I know that recently you've been doing more kind of like clean up and maintenance work, it sounds like. Just, you know, keeping things in order. But I'm curious generally why—tell me a little bit about why you work with AFC? What about that work is important or interesting?

I mean, it kind of gives me the best thing for my buck on time. When I find an article that's very notable already written, all I have to do is click the button and because of me there's an article that should be in Wikipedia, that might not have gotten to Wikipedia either at all, because AFC kind of has an auto-decline culture, or certainly not for a while. So I find that aspect a little bit rewarding.

And how often do you think, would you say, how many of the AFC drafts that you review do you end up accepting?

Certainly less than 1 in 5 , definitely less than 1 in 10. If I was really fixing everything that was fixable, regardless of whether it was a good use of my time to do so, I could probably save about 1 in 4, 1 in 3.

Interesting. That's higher than I expected, but encouraging I suppose.

Now a lot of those are things that just the subject is notable but I would have to essentially re-write the article.

Right. So when you say that AFC has an auto-decline culture, what do you mean by that specifically?

I mean that it's a lot easier for "this might or might not be notable but you haven't provided the sources" rather than to go Google it for 5 minutes and then say "give me some sources", [instead you] hit 'decline' and [say] "sources do not show notability". If it's written in a puff style, an advertising style, it's much easier to say "decline-NPOV" versus go and re-write the whole article and now you have a marginal Wikipedia article that would probably survive an AFD and probably is a net positive for the Encyclopedia. Now that gets taken too far, sometimes. Some of it is good, some of it is just editors using their time appropriately and not spending 3 hours fixing a draft that may or may not be notable. Some of it is editors who should take the 10 minutes to clean up the wording on two paragraphs but instead just hit the decline button. So...

Do you ever go in and try to dispute or re-submit AFC articles that have been declined by other editors? Do you ever contest those decisions in any way or try to get them reversed?

I would if I saw one that was very clear-cut, but usually I think it's a better use of my time to just look at a fresh one. There's always a backlog, there's always fresh ones.

Right, so it's not a matter of trying to salvage the trickle of articles coming in: there's a flood, and that's how you're prioritizing.

And the ones that get declined, maybe 1 in 10 are salvageable if I really want to put the work in. Of the fresh ones, it's maybe 1 in 5, 1 in 4.

How much work, when you say that you've put a little bit of work into articles that you want to accept, what kind of work and how much of it—how much effort are you generally willing to expend for a particular article that you think makes the cut, and what kind of work do you generally put in?

It depends on [not captured]. I'm generally willing to put more work into something that was clearly a good-faith, non-paid, non-advertising—the editor was clearly here to build an encyclopedia article. If you wanted to submit an article on an obscure genus of moths, I'm much more likely to spend time on that article than on "company X that makes widget Y", and other promotional stuff.

I want to dig a little further into that. What are some of the things that you look for when you're making the decision about whether an article is written in good faith, as you put it?

So, whether it's written in good faith? One of the things that I really look for is promotional language and whether it's on a promote-able subject. If it's "company X"—is a big red flag for me. "Professor X", "big person X". Those are all red flags. But if it's something objectively notable—like, one of the most popular AFC creations I've had in terms of edits after it got into main space was on a 16th or 17th century Chinese [not captured], which I thought was an interesting article although it's way outside my field.

Right. What is your field? What are the areas that you like to contribute to as a content contributor?

I don't contribute a ton of content, my educational background is in STEM and I generally stick to those articles if I'm going to be making major content creation or additions. Although not always. I just [nc] an article on a company. So I guess my normal education doesn't necessarily intersect with the WIkipedia articles I edit. I also do a lot of behind the scenes work. Besides AFC, I've been on a closing contentious RFC's kick, which is fun as a non-admin.

There aren't very many of you doing that.

There's like 3 others.

When you're working on AFC—I want to ask a little more, I think I got a little information from you but then I asked a different questions and took us on a left turn. Why is AFC important? Why is it important to do that kind of work?

To be honest, I wish the process didn't exist. I think that if we just had, if you want to make an article on Wikipedia, either allow IPs to do it after their IP had registered the appropriate edits for autoconfirmed etc and allow IPs to gain the privilege I guess. Or else say, "register an account" and get to autoconfirmed. You see a lot of people who have no idea how WIkipedia works trying to submit articles. Why I choose to work on it is because it's got the highest concentration of potentially very good articles and the ones that I think can never be brought to main space acceptability can be declined quickly.

"Highest concentration" versus what? What are other ways that you could review new articles?

I could make articles, but I think I can get content, good content, into main space faster by reviewing other peoples' work than by writing my own. Even though I'm sure that 95% of any articles I made would make the cut.

Have you ever tried the page curation?

Like reviewing?

Yeah, the new pages feed.

I did a little bit of new page patrol work but I would classify that far more under anti-vandalism than under content creation. And ClueBot does so well with that that...

Have you ever used the... I'm going to throw a link into the sidebar... have you ever used the Special page New Pages Feed?

Yes, I've used that before. If I'm doing new page patrol, which I do seldom, that's what I'll use to do it.

When you're looking at New Pages Feed, tell me a little about how your process differs whether you're looking through the AFC backlog and doing reviewing there versus doing reviewing here on the new pages feed.

I'm doing something completely different on those. On AFC I'm actively trying to make the encyclopedia better. On new pages feed I'm trying to stop the encyclopedia from getting worse. Which means that I basically look for the vandalism that ClueBot misses. So I do profile on that. I will be far more likely to click on an IP editor, I'll be far more likely to click on...

A red link? A red linked userpage?

Maybe. Certainly a redlinked talkpage. I'll be far more likely to click on anything that's been tagged, as I should be. And then I'll just kind of click randomly if none of those are popping up and doing that and just click, look at the diff, this was clearly made in good faith, go back to new pages.

So in that case, you're mostly, the work you're doing is much more just point-and-click. Very rapid, vetting, marking problematic things, kind of giving a check mark to things that seem like they're ok for now and moving on. Is that right?

Yes. I mean I won't even read the article I will look at the diff and basically make a judgement only on that about whether I thought this was in good faith. And if it looks like it was, I will just go to the next link.

Why would you say that is? Is it because the signal to noise ratio is so... versus AFC that it makes you want to skim through quickly and get rid of the crap? Or is it something else? Is it because the process, the gadget is designed differently?

The gadgets are definitely designed differently. AFC generally makes bigger changes than New Pages Feed. Although actually, sorry, it's not usually New Pages Feed, it's recent changes that I'll normally use. Which, that's a big difference. Yeah, I almost never go on New Pages.

Okay. Is that just because it's not a type of work that appeals to you? Or is it because you just don't like the interface?

Usually it's... I mean in AFC it kind of feels like you have a little bit more control and guidance. You can edit the page before it gets to main space. Or you can decline it with feedback, although that seldom works. Versus new pages feed, it's like I would have to look at this... there's not really a good tool for it, like there is with AFC Helper. For that, if there was a tool that let me look at this, and then have a drop-down menu that said "tag whatever, blatant/patent nonsense, tag attack", although I've never actually seen one of those.

Now that you're talking about the AFC Helper Script. I have it installed, I've never had the guts to review an article, but I've opened it up before. Can you go to AFC and kind of mock your way through a workflow for selecting an article to review and get to the point where you're examining it and using the gadget?

Sure, I'm going to go to Pending AFC submissions. I typically go to age and I typically like to review fresh ones, because I think that if I can get this back to the editor in a timely fashion even though I'm letting some through the cracks, that draft has a chance to get a second revision. So I'm going to pick one at random, let's look at "Mario Joseph". "Mario Joseph is a former Muslim imam from Karala, converted to Christianity..." and just looking at this I don't see the evidence of notability, I'm going to open up the links which I'm being pretty generous by doing this for this draft. I'm actually seriously considering declining it on notability right now.

Why is that?

Just reading the links, I see an interview with the guy, another interview with the guy, and a third interview on a Catholic podcast. So, I'm also wondering what this actually adds to the encyclopedia.

So it looks like even though there are sources here, they aren't necessarily...notable?

Not what I would call reliable, no. And I mean ultimately the test I'd use is whether this would survive an AFD debate. And I think the answer is no. So I'm going to review it with the AFC Helper, well actually what I'm going to do, and this is something I would love a tool for, I'm going to take a good chunk of this, copy paste, see if Google comes up with something. It doesn't.

Are you looking primarily for more sources, or for copyright violations?

I'm looking for copyright violations. And here I see it's been deleted three times, which is unusual, this doesn't come up very often. I see two blatant copyright infringements. This isn't one of those. And this also looks like it was a different Mario Joseph. So I'm going to go ahead and decline with my standard notability bio, and say "while you do have sources, they are not really reliable" and give the link to Reliable Sources. If I decline, I always keep these boxes checked, because they're good.

The Teahouse invite

The Teahouse invite is one of them, but also the notification, and then it just takes care of everything for me. That said, I'm very pessimistic that this article is ever going to come back in a form that WIkipedia can accept it. But it's one of those that if I really wanted to I probably could have saved. He was probably notable-ish. He was probably on the line. But I would have really dig for sources.

Is that about the level of review commentary that you generally provide when you're doing a decline?

That was probably generous given what I normally do. Just because there are so many and I have to [nc] if I want to... the [nc] is optimized to get the maximum amount of content into Wikipedia while not biting the newbies as much as I can, and while keeping in mind that that could come around again, and be notable. I didn't look for sources, I could have missed the three BBC articles and the CNN article on this guy.

Do article creators ever contact you after you've declined or accepted their articles?

Seldom. 1 or 2 percent.

So even though you're going for, specifically targeting articles where the person who created it is still around, you still don't experience that they reach out to you very often to ask why or ask for help or anything like that?

Yeah, almost never. Which is one of the reasons  that if I seldom receive feedback or someone contacting me to try and help, why should I take the time to write an extensive, you know? Because I could have easily written two paragraphs explaining exactly how to fix that. Instead, I spent 15 seconds on it.

What's the most amount of time you would spend to actually bring an article what you thought was a standard of notability or reliability or wikification, or whatever? If you chose to really take on an article as your personal project to get it into main space.

I would happily spend a couple of hours on a clearly notable topic that's free from puff or potential COI, where the sources are there. For example, an article on the [poker tournament continuation?] that came through. It was rough, but   that's an article that I would probably spend a good chunk of time working on. It's one that I've actually thought about just writing, because it should be there and it's not.

The AFC Helper script seems like it works pretty well. Is there anything about that tool that you wish were different or that doesn't work as well as you'd like?

That script works amazingly. The only feature that it doesn't have that I wish it did was a copyright check. Those exist, if you tag a page as unambiguous copyright infringement there's a little script that will inform the deleting admin to do a copyright check with the website you provide and also every other website on the internet. If we could incorporate that into the AFC helper tool so that you saw how it gave that this was previously deleted? If it also gave that there was a copyright infringement of these websites, you could have a human editor just check that, if a human editor thought it was 20%.

At AFC it sounds like a lot of the stuff you're seeing is promotional, or advertisement, or not reliable sources, so that comes in conflict with notability or COI I imagine. What are the most common easily fixable errors you see on good faith AFC drafts?

Specifically on good faith AFC drafts? There's often very poor use of English language. And I'm going to actually separate this into categories. There are a lot of towns from India and places in India that get submitted to AFC where they're absolutely notable and I can't fix the language. That's about 10% of the drafts I review. Outside of that, 10% in the general category of very poor English, but a clearly notable subject. Outside of that, the remainder of the good faith articles there are a lot of good faith articles where the subject just isn't notable and those just can't be fixed because you can't fix notability. Of the fixable good faith articles, maybe a total of 15% of the ones I review are ones where it's not presented neutrally but it's absolutely notable and they've tried to source it, and those are probably the easiest to fix because all you really have to do is read the sources and re-write the article. Which, that'll take you 40 minutes for a 3-paragraph article with 10 citations or something.

So it's just the language that's inappropriate, it's not being presented in an encyclopedic fashion.

THat's something that commonly happens, yes.

And would you say that generally when you notice those you generally go through the process of trying to clean them up?

No I generally decline them.

Because ultimately it's still 40 minutes of work?

Yeah, it's still that 40 minutes of work. I'll clean it up if it's especially notable or especially close, but if it's pretty far away and fringe notable, no. And that's the majority of them.

So it sounds like a lot of the articles that are getting declined are articles that could be Wikipedia articles, but the person who submitted them doesn't know how to make them Wikipedia articles and the AFC reviewer who's reviewing them doesn't feel like they have the particular subject matter expertise or the language expertise or the familiarity with the sources or frankly just the time or desire to devote to salvaging this particular article. Does that sound relatively accurate?

That sounds close. The thing that I would add to that is that if we suddenly had every AFC editor give the full time that every draft deserves to really try to save it, the backlog would just be 3000 pages in 3 days.

So ultimately the AFC reviews have found an equilibrium point in terms of how much time they can devote per article, balanced against the stream of articles that are coming in to try to keep the backlog from growing at very least.

That's probably fair, yeah. And you saw what happened when I did that, "Mario Joseph" could plausibly have been an article, and instead of declined.

What do you think... it sounds like you are somebody who believes it would be a good thing potentially for the encyclopedia if there were more people who were able to review these and there were people available to actually, who had the ability and the desire to actually work on salvaging some of these. Or a greater number of these articles. So if backlog wasn't an issue, what... maybe I'll change the question: what do you think would need to happen to allow more people to... I'm still not asking the question right. What would need to happen to get more of these borderline articles through AFC? Obviously more people is part of the answer. But what are some other parts of that answer?

More people is definitely part of the answer. If we could find a way to get more of the drafts close or else feel that I can accept that if "Mario Joseph" goes into the mainspace, the main space editors would edit it. If we could change the culture. But how to allow that. It would make it a lot easier but I know that if I just permitted that, it would go to AFD promptly.

So if it wasn't the norm to immediately mark for speedy deletion articles like that when they appear in the main namespace, then you would feel more confident allowing more of these borderline cases in. But you don't want to allow articles in that you don't have time to fix and that you know will not last long in main space.

I don't want to let articles in that I'm not willing to fix, assuming that another WIkipedia editor is going to fix. I don't think "Mario Joseph" is speedy deletion criteria. I do think it would have been PRODed or contested, and so on.

Do you think that there are people on wiki who, if you had known how to find them and come to them today, would have  been happy to take on the work of improving the "Mario Joseph" article? Is there an editor out there that if they know that article was about to be declined, would be likely to come in and work on it?

I doubt it. But if you want to try to go down that line what you could do is notify this WikiProject and I'm sure WikiProject Religion or WikiProject Islam or WikiProject Christianity would have probably liked to know that I declined that. So if that was just a checkbox in the script or a drop-down menu or something, maybe. I'm still not sure it deserves the clutter, but honestly I don't think there's an editor that would care about Mario Joseph other than the one who submitted it.

But there are articles that are coming across and are getting declined where there might be subject matter experts out there who would be interested and might try to get this article boosted into main space.

Absolutely. I could find one right now if you wanted. I don't think it would take long.

No, I believe you. The other thing I wanted to talk to you about is—so we've been talking about article review. Article review is obviously as you've shown, as I've gleaned from other AFC editors I've talked to, it takes a fairly well-tuned human classifier to be able to do article review well. To be able to efficiently triage these draft and really know what to look for in terms of signs of promising article or signs of an article that's a lost cause. Potentially there are ways in which this process could be improved, and some of those might even be making improvements to the AFC Helper script itself, but it doesn't sound like this is something that could be easily automated. In fact, it kind of sounds like the New Pages Feed, which is in some ways an attempt to automate some of the se processes, doesn't really work for saving new article that are potentially worthwhile but need work. It sounds like it works most effectively as a way of rapidly getting rid of a lot of the crap, and some of the borderline cases with that.

In my view that's the purpose of the New Pages Feed, is to get articles out of main space while they still haven't been there a long time. Get the hoaxes out before the newspapers find it. Get the hoazes out before the student finds it. Get the advertisement out before potential consumers find it. [nc] while the people who wrote it can still be apprized of their mistake.

Now I want to talk a little bit more about edit review activities. You said that you sometimes review recent changes. Do you do any other kind of—other than your personal watchlist—edit review or antivandalism type work, with Huggle, Twinkle, AutoWikiBrowser, any of those tools?

No, I don't really use any automated tools. The only one I use on a regular basis is AFCHelper which might kind of account for how magical it seems to me.

It seems like a really good script to me too. It feels kind of magical. Good software sometimes feels kind of magical, considering all the other crap that we have to use on a daily basis that's not magical. Tell me about how often do you check recent changes and do edit review stuff there?

I haven't done it in [18?] months. I will go on a kick occasionally where I will go and do 25, 50 you know "look at random edit that I [nc] based on profiling or whatever, undo it if it's not good faith, move onto the next one" but it's not something I do often.

Okay in that case I'm not sure that this is the most interesting line of discussion that we could go on. So, to give you some background: one of the things that, the possibilities that the Collaboration team is pursuing is to build some machine-learning improvements into a RecentChanges-type feed, that would show only edits by new editors, and a particularly only edits by new editors that have a high likelihood of being good faith, but somehow including an error or needing work. So it could be that a substantial amount of content was added to an article that doesn't seem to be blatant vandalism, but that clearly needs sources. Or it could be that somebody edited a page and it inadvertently broke some markup in the process. So one of the tool ideas that the Collaboration Team is thinking around is to create a feed of just these types of edits that have been pre-classified as good faith, but needing help. And presenting that as a way for editors to, experienced editors to intervene, fix some of these errors, but also to give people an opportunity to educate these good-faith new users rather than treating them like vandals.

I can tell you exactly what's going to happen with that. You're going to have a couple of editors who follow this page and go "new editor problem-->appropriate tag slapped on talkapage". And you know, those tags if those are your first experience on Wikipedia, good luck! I was lucky. I got adopted in the Adopt-a-User program very early on. But that was a 20 hour commitment by an experienced user who is now an Admin. That he could have been improving the encyclopedia in other areas [during that time.]

Right. So a system like this is very open to just being used as another way to more efficiently revert & warn, it sounds like.

You're going to have interest with people biting the newbies, yeah.

So if this tool is meant to, if we're trying to build something that makes that less of a problem, this probably isn't the right way to do it. Is that what you're saying? If we don't want newbies to be bitten, we shouldn't do it this way?

I'm saying if you want to do it this way, you really need to filter which users are viewing this feed. Yeah, that's not possible. Wikipedia will self-select in the editors and... it's a decent idea. I also am nervous about whether it's just going to be extraneous, whether the tools we already have like new pages and recentChanges do it well enough that we don't need it.

What do you think is a good way to support new editors who have, maybe they just created an account today, they've made a few edits and they're bumbling around and making rookie mistakes. What do you think is a good way to support them?

That depends on how much time you have. If you have unlimited time? The Adopt-a-User program would be wonderful. But we don't have experienced editor time; backlogs are growing as we speak. If you've got limited time, the best way to do it is a tag accompanied by a personal message that obviously didn't come from a robot. So if I was given 60 seconds a page and told to correct the newbies without biting, I would apply the appropriate tag and then write a personalized message that [nc]. If I'm only given 5 seconds a page, I have to tag and hope that the tags are written well enough. Any improvements that you make to a commonly used new user warning tag will pay dividends if it's transcluded often enough.

Although making those changes is generally controversial.

I mean if the WMF does it, then yeah maybe.

Yes. Correct.

That said, if you leave the suggestion on the talkpage of the template, possibly it has a chance of passing. You know, it could get in there and do some real good. That said I think most of those user warning templates are for what I've seen pretty decent as-is. I don't know how you make them much better easily.

They're pretty straightforward and they may come off as stern but ultimately they're not aggressively worded just very statement-of-fact?

I mean the level 0 ones aren't even accusatory at all, it's just "be aware of this" or something. Another thing that I think is very useful is the Teahouse. When it gets used right it generally from what I've seen is pretty good results. So promoting that a little might be nice.

I'm one of the people who created the Teahouse.

Good job.

It took off better than we ever expected. We will be promoting the Teahouse more. We're trying to, working right now to get a Teahouse set up on Swedish Wikipedia. And I've been talking to some Teahouse editors and it's amazing how much folks who are answering questions at the Teahouse really care about giving good quality answers and being supportive. Do you answer at the Teahouse very often?

I might have done it once or twice. I just lurk there occasionally and see the good results.

Would you be interested in editing at the Teahouse? Is there anything that would make it more desirable to you?

I haven't seriously looked at it so, I wouldn't know. The final thing that I could think of for a new editor would be make it so that they have an experienced editor who they think they can turn to and just ask the questions without you know biting. I found it pretty quickly that it was actually the user who adopted me, before he was an admin, adopted me and I could go and ask a question if I had one. I didn't have to actually use it that often but even a message that says "hey, you look like a new editor, if you have questions, talk to this person, he's an experienced editor he can help you", and then just very carefully screen who's in that.

So even if it's not a full-scale mentorship relationship, it's still useful to have a touchpoint, someone you know you can go to. Yeah, that makes sense.

I have no idea how you would do that. The Teahouse is the closest approximation of anything implemented that I can think of.

You've been editing for at least 4 years now, 5 years now. As a registered editor, on and off. Do you feel like the culture of Wikipedia has changed in that time?

Immensely. It's so much more rules and policy focused. When was the last time you saw a successful invocation of IAR?

Yeah, it's kind of a joke now.

I mean, IAR is used when a policy is poorly written, but if you're a new editor walking in I have no idea how you would do it. It's like "no this draft is bad because of these 5 three or four letter acronyms, you proposal is not going to be accepted because of WP:SNOW..." and, you know.

Not exactly the most heartening thing to say to somebody who just showed up and spent a couple hours dealing with markup and trying to make a contribution to the world's largest encyclopedia.

And you reject them with three letter acronyms.

Are there things that you can think of, or strong suggestions that you have that might help newcomers in... I'm just looking for ideas because I want to go back to this team that make recommendations that have them focus their energy in a productive area and if the goal of their work right now is finding new ways of helping experienced editors support good-faith newcomers, it sounds like edit review might not be the right place to be doing that. Any other ideas you can share?

A small-scale idea would be to see if we can make it so that instead of linking the neutral point of view policy as WP:NPOV, if we could link that as "Neutral Point of View", that would be a little less offputting I think. On a medium to long term scale, if you make something that can fit on an 8.5x11 piece of paper when it's printed out sheet of "follow this advice and you will not get yelled at by editors". You know, just summarize every policy that you need to know in.

Do you think it makes sense to discourage new editors from trying to make articles, or drafts?

I think that if you put it up for community consensus that they would say that new editors still should be creating them, although I think it would be a discussion to have. The Siegenthaler Incident was, what, a decade ago?

I'm thinking more from a newcomer survival point of view. It sounds like starting up and trying to create an article before you know all the pitfalls seldom leads to a good outcome.

You're going to get rejected 5 times before you get it accepted. Which leads me to another problem, is that as an AFC editor you're sometimes between a rock and a hard place. Let's say you have a draft. Even if it's potentially notable, it fails on 5 different policies. Do you give them a decline rationale with 5 things that need to be fixed or do you decline the most obvious thing, and move onto the next article. I think it's better for the new editor if we give them those 5 and say "hey, here's the Teahouse, or ask me a question anywhere you like", but that's not going to happen and generally what happens there is I cut my losses and decline on the most obvious or hard to fix one of those, and then I've put that new editor in a bad spot because if they fix that they then get declined for the next most obvious reason, the next reason up. But I've also not wasted time on a draft that I think had no shot.

And ultimately if they're willing to put in the effort to fix that one thing, there's a reasonable probability that they'd put in the effort to fix several other slightly less pressing issues.

Maybe, maybe not. They might just think that Wikipedia is this labyrinthine bureaucracy which, to be fair, they're not wrong, and give up on it. Which I've seen happen repeatedly.

Alright, personal connections with actual people and individual support some way or another seems like it's the most effective way to keep somebody holding on while they learn the ropes.

I would think so, definitely. Adopt-a-User is obviously the Cadillac program for that.

Is that still working? I notice that you have somebody adopted on your userpage right now. Is Adopt-a-User still relatively active?

I have no idea. I suspect it happens for a dozen users or so every month or year or something but I haven't checked. Something to check.