Talk:Wikimedia Fellowships/Project Ideas/Editor engagement strategies for new or smaller Wikipedias

Research and Experimentation

This feels like a really big project as designed, and I expect success would be found by choosing carefully what to take on and what to leave out (for future projects to do, of course!) Phases 2&3 might be most manageable if the set of small wikis chosen for experimentation was very small. I'm picturing a handful of languages at most to start, and just a couple of different test areas to focus on (based on what you'd learn from phase 1 research, say, just testing welcoming only, or x number of approaches to welcoming). I wonder if this is how you're thinking of it too? And in that case, maybe phase 1's research aim would be focused on learning just enough about small wikis in general to determine which wikis to experiment on and what to prioritize for experimentation? (This is, I think, similar to how you're thinking of it, already, but its always worth being clear about the focus!) What we've learned from past community research projects is that there is a LOT to learn about even just one language community, so having good focus from the beginning is really important. Siko (talk) 01:00, 25 February 2012 (UTC)

I will answer part by part. At first, just to clear the plan again, if you think all the phases include all small Wikipedias, then no. Only phase 1 is the phase where we will work on all possible small wikis to collect data and find out usable information and scenario to process and plan the actual possibility of running the next phases in those Wikipedias. And based on what we will have after processing, we will select a set (or few sets, depending on the plan) of Wikipedias to work on.

This feels like a really big project as designed, and I expect success would be found by choosing carefully what to take on and what to leave out (for future projects to do, of course!) Phases 2&3 might be most manageable if the set of small wikis chosen for experimentation was very small.

Yes, it is a big project, but I am not actually thinking about leaving something out, not even for future projects. Alternatively you can (maybe) think about a group of phases as a research area or separate projects. As I said, phases 2 and 3 will be chosen only for a set (or few sets) of small Wikipedias based on what we will get from phase 1.

If you really think about parting this project, then you can think about phase 1 a separate project. But that includes only research and field work. We will just gather information on small wikis. Those information will not be very useful if we don't step to the phase 2 and 3 quickly as I planned. Phase 5 can be a separate project too, but well, that is a small phase, and it's on implementing, what is about doing something almost full-prof. So it can be arranged with this project too IMO.

I'm picturing a handful of languages at most to start, and just a couple of different test areas to focus on (based on what you'd learn from phase 1 research, say, just testing welcoming only, or x number of approaches to welcoming). I wonder if this is how you're thinking of it too?

We will start with a handful of languages (as I cleared above). We could test the types of welcoming only, but that is not my plan actually. Because lack of a good interactive welcome message is not the only of the big needs. There are some other needs, like lack of handful help pages and policy pages to explain the basic policies and Wikipedia's manual of styles of writing. A page to ask help easily, etc. If we test only the welcoming method at a time, we will get data on how much users got that message, and maybe how responsive they were, but that will not help with editor retention at the end. Because if the links of the message are not helpful, do not provided the answers of the questions of the new editors, and if that wiki does not have a friendly way to ask help, or does not have enough users to mentor or help or welcome the new ones, the whole welcome test will not give much success. Welcome messages are more like a light to the entrance, but constructive helping system, mentors, necessary help/policy pages are the light to guide a new user all the way to the end and to make him/her an effective editor.

Or we can use few sets of Wikipedias. One will go with automatic welcome, one with manual, etc. In my view, this is the only thing we can take optionally right now. Like, help pages, policy pages are compulsory, if a wiki does not have them, we need to create them anyway to work there. Or a welcome message will be a message with unhelpful links. But a lot to plan about that after phase 1. Remember outreach could be an issue too. We may need to make a way for that in few projects.

And in that case, maybe phase 1's research aim would be focused on learning just enough about small wikis in general to determine which wikis to experiment on and what to prioritize for experimentation? (This is, I think, similar to how you're thinking of it, already, but its always worth being clear about the focus!) What we've learned from past community research projects is that there is a LOT to learn about even just one language community, so having good focus from the beginning is really important.

Exactly, that is something I am thinking about. Phase 1 is just about knowing the small wikis. After knowing we will process the collected information, and determine what projects we should work on primarily and also to work on what (early step of phase 2). There is a lot to know, and I am sure we will know them in the process. :-) — [ Tanvir | Talk ] 10:47, 26 February 2012 (UTC)

Measurement and Implementation

Being able to iterate quickly on what you learn from each experiment feels really crucial for this project. Are you picturing doing some of the measurement and reporting during the earlier phases or saving it all until the end of the project? I think its an important and great idea to build measurement, reporting, and an open call to get other wikis involved into the timeline! But I might encourage you make sure you'd be building some metrics and reporting into the process early too. Siko (talk) 01:00, 25 February 2012 (UTC)

There will be reports in the middle of the processes, hopefully you will get report after completion of every phase. And of course, the result will be on-wiki. The data, the feedback, we will get from those small communities will be on Meta, so other can do analysis on that, and give us the feed-back to make this project successful.

There are some metrics in phase 1 to know the present status of a community. But I am not sure what do you mean with "metrics and reporting into the process early". I am clear about the reporting, but not the metrics. We can do day-to-day analysis to keep an eye on the changes will be made in phase 3 though. — [ Tanvir | Talk ] 10:47, 26 February 2012 (UTC)

Phase 2: Research

Phase 2 talks about designing welcome messages for _all_ small Wikipedias. I have a problem with this, since I think adopting a one-size-fits-all method is not the most effective thing to do. It is, of course, important to build fancy welcome messages, but what if the people from that particular region don't know whether that certain small Wikipedia exists in the first place? I believe that we should look for strategies to make these smaller Wikipedias more noticeable whilst(or before?) building welcome messages, otherwise we might end up with a Wikipedia with a beautiful welcome door but no visitors. Abbasjnr (talk) 15:43, 14 March 2012 (UTC)

Thanks for your feedback Abbas. Actually, that welcome message will be a general welcome message which will fit all the basic needs. Considering the differences between projects, we can (during the pilot projects) or for sure they (respective communities) can change it themselves. I think you will also agree here, we cannot think about every particular needs of each small Wikipedia at this moment. So we will be particular and thorough only to those projects where we plan to run the pilot. Others will meet the basic general needs which are same in all small Wikipedias.

It is a good question, what if the visitors don't even know about our efforts of welcoming them? That is why we are thinking about running community driven on-line/off-line outreach program (with our support if needed) in those areas to let editors/people know. You can see that I have emphasized on outreach program in my plan (see the third point under “Following steps should be done for those Wikipedias on pilot projects” section. Also, to let them know about this we can also use site notice (or central notice) system. I hope these make sense to you. Regards, — [ Tanvir | Talk ] 16:53, 14 March 2012 (UTC)

Thanks for the response! Abbasjnr (talk) 05:10, 22 March 2012 (UTC)

Selected for fellowship

Congratulations to Wikitanvir, who has been selected for a Wikimedia Community Fellowship to work on this topic area. Tanvir's work is beginning with Bangla already, and anyone interested can follow this project on Tanvir's meta project page, as well as on Bangla Wikipedia. Siko Bouterse (WMF) (talk) 21:39, 30 May 2012 (UTC)