@groceryheist @mako you're the only ones I can ask. Here's a plot of Chinese Wikipedia daily editors. The story here, of a community that overcame exponential growth and found stable linear growth, is mirrored in the daily edits curve (and to a lesser extent, the new articles curve; not so much the bytes added, that's been flat for many years).
My question: is this story real? Are they actually mentoring editors, growing quality, and expanding? Or is this just PRC vs ROC turf wars, or spam?
@22 Thanks for your great questions!
I haven't looked closely at Chinese Wikipedia so I can't speak to it directly. It's very interesting that they have stable linear growth. And I would like to know more.
One important fact that doesn't get emphasized in the paper is that there is a /lot/ of variation between wikis. The average trend is consistent with RAD dynamics, but I wouldn't claim it is universal.
@mako anythinng you want to add?
@groceryheist @mako I'm burning all my mana on my Chinese-speaking friends to get me some insights, since I'm still at least five years away from answering them myself 😅, hopefully y’all will have more luck.
I dig what you say about variability on Wikia. My hypothesis, having looked at just a few nation-grade Wikipedias, that stable linear growth is what I expect to see (new articles created, old articles freshened), and that other, smaller, Wikipedias may have that, rather than RAD. Will look.
@groceryheist Sorry, I don't follow (insomnia)—how do you mean 'less likely to generalize'?
(I'll say this here because I've been thinking about it and have nowhere else to say it—I learned how to ask questions and answer them on StackOverflow by people (not so) gently telling me when I was doing it wrong, and that (rough) mentorship is something I wish Wikipedia had more of. (I kind of alluded to this in my initial question—are Chinese editors mentoring newcomers, thus keeping them?))
Oh! I'm so sorry for not being clear. I meant to say that we chose to analyze Wikia rather than Wikipedia editions because, from the perspective of replicating results from EN:WP, we wanted to see if they would hold up on communities that /were not/ Wikipedia. Like maybe there is something weird about Wikipedia that leads to rise-and-decline dynamics, but that would not happen on Wikia. Does that make sense?
@mako @22
As far as mentoship goes, I haven't studied that deeply, so I can't speak to that very much. My intuition is that whether a given style of mentorship would work well depends a lot on the particular newcomer and their personality, motivations, and their prior experiences online. In theory I would say communities should try to intentionally pair mentees with mentors that would be a good fit. Unfortunately, I have hard time seeing how that would be doable in practice.
Thanks for weighing in on this unusual topic too! I have a more fluid, serendipitous, unstructured view of "mentorship"—the kind I received and try to provide on StackOverflow, GitHub, etc. Any community member with more experience than me could give a drive-by comment on my question/answer (StackOverflow) or pull request/issue/repo (GitHub) and I learned "the right way". On Wikipedia, my experience has been unceremonious reversion of my edits, no explanation.
My experience informs my hypothesis that communities that see stable growth must have some way of grooming beginner members. StackOverflow actually has spent a ton of effort tweaking the mechanisms to transmit/receive quality signals to/from the community. I don't think Wikipedia has a great feedback story of any sort, so I'd be curious how Wiki-communities that do give feedback (if any exist) do so—extensive use of talk pages? IRL edit parties?
@groceryheist @mako Thanks for listening! Y'all are the only ones I can ask about this :)
Aha, got it! Here's my sophomoric thinking—RAD is what I'd expect in most human endeavors, from Pokémon to the Grand Canal, and I'd expect to see it in abundance on Wikia. But nation-grade Wikipedias seem…different. I'd expect a healthy Wikipedia to have linear growth as it (1) creates new articles and (2) freshens old ones. RAD there means the speakers of the language don't care about the Wiki principle/goal (& there are interesting reasons why that might happen).
But I'm also really curious if there are big Wikipedias with nice stable ~linear growth. So far Chinese is the only one I've found (I just randomly picked 7 languages for initial analysis), and rather than poison myself with data, I'll set up a proper experiment and see how many other Wikipedias may have linear growth. Then I'd want to see if that linear growth did indeed indicate a vibrant community or if the explanation was seedy (spam, edit wars, etc.).
@groceryheist @mako @22 i'd be kind of curious what results would be like (if the data even existed at this point) for the early days of the wiki form - c2, usemod, meatballwiki - or adjacent ideas like everything2.
@groceryheist some of that era's projects still online, at least in archival form. c2 (ward's wiki), for example:
and usemod / meatball:
i'm not sure how accessible the history is at this point.
@22 @mako We chose Wikia in part because it seemed less likely to generalize. But your intuition is the opposite! We should really do an iteration of the analysis across different editions of Wikipedia.