From Meta, a Wikimedia project coordination wiki

Village Pump discussion on English Wikipedia[edit]

Please see this Village Pump proposal currently under discussion:

One of its purposes is to help stabilize the number of active editors. --Timeshifter (talk) 09:48, 27 July 2012 (UTC)[reply]

en:Wikipedia:Village pump (proposals)/Archive 91#New class of admin-moderators to resolve content disputes.
See also:
en:User:Timeshifter/More articles and less editors. --Timeshifter (talk) 15:36, 21 August 2012 (UTC)[reply]

What the page/team is about[edit]

The page doesn't currently explain what "editor engagement" means/what's the background, it should probably link to the strategy wiki pages as there doesn't seem to be any on Meta for the purpose. See Editor engagement. --Nemo 05:23, 21 August 2012 (UTC)[reply]

en:WP:E3 and en:Wikipedia:Editor engagement have more in-depth background material and explanation for the overall strategy. It's there instead of on Meta because most new WMF features and experiments use English Wikipedia as a proving ground. Right now meta is used primarily for research and data analysis documentation. Steven Walling (WMF) • talk 05:59, 21 August 2012 (UTC)[reply]

Something about testing enwp's hilariously bad editnotices[edit]

Could perhaps the effectiveness of horrible garish editnotices versus not so horrible garish versions versus nothing at all be tested on enwp, since they use a lot of those? The editnotice for WP:SIG, for instance, is a prime example of how horrible and garish they can get, and it was put there to try to keep people from asking dumb questions, but of course some still do... so the question is, how well does that sort of thing actually work, and is it really an improvement over something shorter, simpler, and with more normal-sized text and styling? Because maybe they really could all be made less horrific and more consistent without detrimenting effectiveness! That would be cool.

You can find a list of page-specific ones here, since different pages have different ones. Many aren't that bad, but others are just... whuuut? -— Isarra 22:21, 20 September 2012 (UTC)[reply]

Convert the IP editors[edit]

There is a large number of goodfaith constructive editors who does not insist to have an account because of certain prejudices. When an IP is editing a page, all he gets is a 'warning':

The message delivered by the above is editing Wikipedia will make your IP address publicly visible unless you get an account. The other benefits link to a page listing the benefits that the user gets (such as the user *ranks*). That page does not tell them how it is going to help the community and the Wikipedia. (e.g., these longterm constructive IP editors wastes a lot of time of our vandalism fighters.) w:en:User talk:Jimbo Wales/Archive 125#Banning IP editing has certainly proved that "there is no reason why a person cannot register if they wish to edit on Wikipedia" and I think the only reason why they are not doing it is the poor edit intro. Remember that editing by anons are done more as a charity than a hobby. So if they are convinced that creating an account is going to bring help to all, then they will surely do it.···Vanischenu「mc|Talk」 22:38, 17 April 2013 (UTC)[reply]

"So if they are convinced that creating an account is going to bring help to all, then they will surely do it."

This is not the case for me. I have no interest in creating an account. But if you think that changing en:MediaWiki:Anoneditwarning or en:Wikipedia:Why create an account? will get more people registerd, then by all means suggest the changes or be bold and make the changes you think will help. Cheers. 05:20, 17 May 2013 (UTC)[reply]
There are certainly some garish versions of MediaWiki:Anoneditwarning. :-) s:en:MediaWiki:Anoneditwarning has always struck me as pretty bad, but as I recall, there are much worse. --MZMcBride (talk) 15:58, 17 May 2013 (UTC)[reply]
That's a good one. 00:15, 18 May 2013 (UTC)[reply]
By the way, I've proposed we add a campaign name to the link, so we can figure out how many anonymous editors actually sign up via that message. Steven Walling (WMF) • talk 01:04, 19 June 2013 (UTC)[reply]

Thank you all for your helpful feedback. I am fascinated to see Research:Anonymous editor acquisition; thanks to Steven Walling and the rest of the WMF for starting it. One million edits a month is very large. We should remove the misconceptions, tell them that all that is required is a username and a password, and bring it to a situation where only those who have technical problems or personal concerns will edit as IP other than the vandals. Think about it. Benefit to all — the editors themselves, the Recent Changes patrollers and the community as a whole.···Vanischenu「mc|Talk」 02:00, 28 October 2013 (UTC)[reply]

@Vanischenu: Hey there. I just thought you might like to know about our upcoming experiment to ask anonymous editors to sign up. Steven Walling (WMF) • talk 22:46, 30 April 2014 (UTC)[reply]

Research idea[edit]

Thought you could try to experiment by welcoming people with a link to WP:WELCOME rather than a template; by warning them against vandalism manually or with template without picture or with full template. I think usage of templates at talk pages is a substantial issue with editor retention at English Wikipedia. --Gryllida 15:01, 8 February 2014 (UTC)[reply]

Overall outcome[edit]

This team existed from 2012 to 2014 and pioneered some interesting work. Its quarterly reviews contain some discussion of what was achieved and (based on the experience gained) considered possible for future such work, but I don't recall seeing an overall summary of stating which of these experiments actually showed impact. Therefore I think it's worthwhile recording here an (impromptu) retrospective assessment that Aaron made some months ago (in response to a question I asked) at a Research Showcase presentation. Transcribed from the video recording:

"[The] strategies we employed on the Growth team to try and increase editor engagement ... were almost all focused at the first few edits an editor would make with the hypotheses that if we can get more people to make that first edit and we can direct them to the right places to make that edit then they will be more likely to have a positive experience.
Regretfully, the only significant results that we were able to show for any of the experiments was either something negative - which was usually when we made a mistake in the interface and made it hard to do something that people generally do - , or a very small positive insignificant increase in the proportion of editors who make their first edit. So this was usually on the scale of 1 or 2% increases, and they were with the experiments around GettingStarted [...]. We saw nothing for any of the metrics that I would refer to as retention. So: short term engagement, but not even what I would refer to as short term retention."

I think one might want to keep in mind the small caveat that few or none of these experiments had resources devoted to studying their effects on longer-term retention. (The aforementioned presentation was about Research:Teahouse long term new editor retention, and stated that this was the first time that a WMF project focused on newbies had been demonstrated to have a significant positive effect on longer-term retention. Interestingly, this exciting finding came only in 2015, several years after the start of that project in 2012, as researchers went back to it and launched a new study of its effects.)

Regards, Tbayer (WMF) (talk) 14:04, 4 April 2016 (UTC)[reply]

Thanks for capturing this Tbayer (WMF). I think it's a fair critique that we didn't really look at long term retention with many of the Growth experiments. It might be worthwhile to run new analyses to look at long term retention on the users who were bucketed in those studies to see if there were long-term effects we missed. I don't see myself having time for that in the near future, but I'd like to support anyone else who'd like to look at this. I'm also happy to help answer any questions. --Halfak (WMF) (talk) 16:00, 5 April 2016 (UTC)[reply]