Topic on Talk:Growth/Personalized first day/Structured tasks

Jump to navigation Jump to search
MMiller (WMF) (talkcontribs)

In this section, we think about two ways to design structured tasks. One teaches newcomers to use the editor to complete tasks. The other is a design that might generate more edits on mobile devices, by making it possible to accomplish a task without using the editor. Do you think designs should be more geared toward teaching newcomers to use the existing editor, or be more geared toward newcomers being able to do easy edits at higher volume?

Zoozaz1 (talkcontribs)

Personally, I think structured tasks should be aimed toward teaching newcomers to use traditional tools. The goal of something like this, in my opinion, should be to ultimately reverse the declining editors numbers and making long-lasting contributions to Wikipedia through the introduction of new editors. The idea of gamification is interesting and useful overall, but I think should be used as an introduction, if necessary a somewhat long one, to fully contributing to Wikipedia. To that end, this project does seem to be a very good introduction.

In the example on the page about adding hyperlinks, I think the first is better suited to the goals of the project. Unambiguous hyperlink additions, if able to be identified by a bot, might as well be added by the bot as well. Ambiguous hyperlink additions would likely need experienced users to go over them. There is a smaller place in the middle of those two where newcomers would be useful, but I still see the main purpose of a structured task as to introduce a user to editing and to make more impact contributions in the future, in line with the first example in the page. The main goal of the project I don't think should be adding fairly inconsequential, in the long run, hyperlinks or other menial tasks (which are necessary, yes, but less important than more complex tasks like article writing). To some extant, tasks like these are a means to an end; yes, adding hyperlinks and other tasks like are useful, but the ultimate goal should be to engage new users with the entire Wikipedia community and having a gentle and frictionless introduction to doing so, so in the end new users can easily transition to working on more complex tasks.

(As a side note, WikiLoop Battlefield seems similar to this idea for vandalism).

MMiller (WMF) (talkcontribs)

Thanks for writing this out, @Zoozaz1! It sounds like you have a strong grasp of how we're thinking about this and the trade-offs involved. I think you're exactly right that we want structured tasks to be an "on-ramp" to users doing more valuable edits, and the trick is figuring out how steep to make the ramp. Structured tasks may accomplish this in multiple ways: (a) it might be that structured tasks teach the traditional tools and so then a newcomer starts using those, (b) or it might be that structured tasks allow a newcomer to very quickly make a productive edit and feel accomplished, making them want stick around and learn more.

That said, we're also keeping in mind that there may be people out there who would make a habit of contributing to Wikipedia if they could find a niche that comes in at a lower skill level than traditional tools. For instance, perhaps there are people who would love to do many simple tasks per day while holding their phone in one hand during their train commute. It would be great if structured tasks also helped those people fit into Wikipedia. Does that make sense?

Regarding the link algorithm, I wrote about this a bit in this comment. Basically, we think that the algorithm we're looking at now isn't strong enough to add the links on its own, but is smart enough to make likely good suggestions to newcomers.

NickK (talkcontribs)

I would say something in between. Letting newbies make a lot of useful edits is helpful, but it should also bring newbies to more casual editing. E.g. after adding first 10 links a newbie should be capable of adding next links at least using VisualEditor, but may choose to continue with this tool.

MMiller (WMF) (talkcontribs)

@NickK -- this concept of "stepping users up" to increasingly more challenging edits is one we definitely need to think about. We need to be able to say something like, "You've been doing a great job adding links. We think you're ready to try adding references. But you are also welcome to keep adding links."

NickK (talkcontribs)

@MMiller (WMF): I would rather say that after adding 10 links with hints we would encourage a user to add links without hints. Simply because hints are not available in real-life Wikipedia editing.

MMiller (WMF) (talkcontribs)

@NickK -- good point. It sounds like you're saying that there two parts to the structured tasks: there's the part where we suggest specific edits, and the part where we guide the user through it. And perhaps after being guided a couple times, a user may only want to get the suggestions, and not have to see the guidance anymore. I think that's something that our team's designer, @RHo (WMF), is thinking about, in case she wants to speak to it.

Barkeep49 (talkcontribs)

What does the Growth team know about how much it takes to have someone stick around? In other words, when a new editor gets to X edits they are y% more likely to get to Z edits? Every edit that they perform that improves something is good for the project. But even better is getting someone to stick around longterm, so I wonder how we can get them to the point where they're "hooked" enough to want to continue even as they transition to "real editing". I put it in quotes because adding a citation or adding an image is real editing even if a person has had their hand held in doing it; When I upload an image to commons using the wizard I am really uploading an image after all.

MMiller (WMF) (talkcontribs)

Hi @Barkeep49 -- thank you for checking out this project! I think you're exactly right with your main point that when we ask newcomers to edit, we have two purposes: (a) getting useful edits done, and (b) investing in the newcomer's potential to edit more in the future.

The Growth team thinks about several useful numbers to guide our work:

  • About 90% of accounts who ever make an edit make their first edit on the day their account is created. In other words, it's not common for someone to create an account and then read and peruse, and then edit on some later date after watching. Rather, accounts that end up editing tend to edit right away. This taught us that if someone doesn't edit on their first day, they probably won't edit ever. So we think we want people to edit quickly, and for that to be a good experience.
  • Depending on the wiki, between 20% and 40% of new accounts make an edit ever (in English Wikipedia, it's 29%). In other words, most new accounts don't even make one edit -- it shows there is a lot of room to grow on getting new accounts to try editing.
  • Between 2% and 6% of new accounts end up being retained, meaning they edit on more than one day (in English Wikipedia, it's 4%). There is also clearly a lot of room for improvement here, since it means that most people who make an edit don't want to come back and do it again.

Regarding the number you asked about, that would be a really interesting one to know: is there a "critical mass" number of edits where a newcomer is more likely to stick around? We don't have recent calculations on that, but that question is exactly what was addressed in a 2009 research paper: Wikipedians Are Born, Not Made. Essentially, the paper says that the people who end up becoming committed editors are usually the ones who do a lot of edits in their first couple days on the wiki. I think the implication for our work is that if we make sure those enthusiastic newcomers have lots of opportunities to edit in their first couple days and have a good experience doing it, they'll be more likely to become long-term Wikipedians. The paper has specific numbers if you're interested, but they are 11 years old, so I'm not sure how close they are to today's reality.

But in summary, I think the way we're thinking about it is: let's focus on getting people to their first day of successful edits, and then to their second day of successful edits. Once we succeed there, we'll be able to start thinking about what comes next for newcomers. Does that sound like a good way to think about it?

Galendalia (talkcontribs)

I think step ups are a great idea. It is always great to build on the knowledge you have gained and to move up to the next level. That would also instill confidence in the new editor that they are doing good or having their mentor point out something not right and where to seek answers and let the editor work on fixing the issue. I much prefer the editor over the visual editor. I find it much easier to use, but everyone has a preference but learning to utilize all the tools in steps will give the editor unlimited potential as time goes on.

Reply to "Teaching vs. volume"