Grants talk:IdeaLab/Email inactive (formerly active) users with a survey on inactivity reasons

Latest comment: 5 years ago by Slowking4 in topic Comments from I JethroBT (WMF)

The obvious problem edit

Why would a user that has left the project be willing to use time on a questionnaire? Only good reason I know of is to be able to identify the culprits that is the reason for the user leaving… Even if that can help admins to take action, I don't know if it is really useful. It might trigger fierce internal fighting. — Jeblad 22:19, 14 July 2018 (UTC)Reply

@Jeblad: There have been a number of people who have left Wikimedia projects or retired who have left accounts or details explaining their motivation to leave. For that reason, I expect some people who have become inactive would respond to this kind of survey. I think there is value in this survey to have more concrete information on how many people leave due to personal or off-wiki circumstances (e.g. job changes, family obligations) vs. on-wiki circumstances (e.g. patterns of bad experience or interactions). I don't necessarily think this has to be about naming "bad people" per se, but perhaps identify processes or spaces that are broken or otherwise need improvement so that they function. I JethroBT (WMF) (talk) 15:35, 19 July 2018 (UTC)Reply
@I JethroBT (WMF): Perhaps a DarnSimple™ questionnaire that only identify processes or spaces… and allow some users to investigate the reporting users interaction with those processes and spaces? Still I wonder if this can be dangerous. — Jeblad 16:38, 19 July 2018 (UTC)Reply
Based on previous surveys like this, many people are perfectly willing to tell you why they quit editing and/or that they haven't truly quit (e.g., editing while logged out, busy with other things but planning to come back, etc.). WhatamIdoing (talk) 16:11, 20 July 2018 (UTC)Reply
imho, naming problematic behaviours (and users) is part of the useful information, all the best, --Ghilt (talk) 15:06, 17 August 2018 (UTC)Reply

It's been done edit

See also Former Contributors Survey Results (completed) and Research:Formerly active editors survey (draft to do it again). User:Piotrus, did you get as far with Research:Why editors reduce activity as you had wanted to? WhatamIdoing (talk) 16:09, 20 July 2018 (UTC)Reply

@WhatamIdoing: After few annoying years of peer review lottery, the paper has finally been accepted and will be coming out shortly in w:Research in Social Movements, Conflicts and Change. If anyone needs a copy earlier, send me an email. --Piotrus (talk) 09:11, 24 July 2018 (UTC)Reply
Congratulations, Piotrus. WhatamIdoing (talk) 15:10, 25 July 2018 (UTC)Reply
Hi Piotrus, i'd love to get a copy, if possible, all the best, --Ghilt (talk) 15:04, 17 August 2018 (UTC)Reply

How would the results be used edit

@Ghilt: This idea has good support from others contributors and there is also past success in administering a similar survey in the past. So, it seems like a potentially useful method to gather information. I'm wondering how the results would be used? Who would want the results and final analysis? And have you thought about best ways to disseminate it to people who would be interested in using the information? SPoore (WMF) (talk) , Trust and Safety Specialist, Community health initiative (talk) 14:51, 17 August 2018 (UTC)Reply

Hi SPoore, e.g. to be posted anywhere and linked to on the arbcom discussion pages and the administrators notice boards? Depending on the sensitivity of the information it could instead be sent to an arbcom mailing list. All the best, --Ghilt (talk) 15:02, 17 August 2018 (UTC)Reply

Comments from I JethroBT (WMF) edit

Hi Ghilt, and thanks for this submission. In concept, I think this idea is sensible and could provide some useful information related to inactivity. I'd like to see how I can help you begin working on this idea, if possible. Here are a few comments/questions:

  • What kind of support or help would you need to start working? You noted in the idea that you'd need a way to contact inactive contributors to take the survey and someone who can do analysis. Do you think there are volunteers who could perform these tasks? Would grant funding (say, through a Rapid Grant) for this work be useful?
  • Have you considered contacting people on German Wikipedia, or contacting some of the individuals who have endorsed this idea to see if they are able to help (e.g. with reviewing survey questions, helping to send out e-mails, conducting analysis?)
  • Regarding contacting inactive contributors, my first thought is that if they have an "e-mail this user" option as a part of their account, or have a e-mail listed on their userpage, that seems like the one viable way to contact them. A second option would be their user talk page, but this is probably a last resort seeing as they are inactive. I'd like to ask some fellow staff members to see if there are any other ways we could consider here.

Thanks, I JethroBT (WMF) (talk) 18:42, 6 September 2018 (UTC)Reply

Well i can ping the endorsers, which i presume are editors on en.wp (except for ProtoplasmaKid and me). In the order of appearance: @Mr. Guye, Hibernian, Omegatron, Winged Blades of Godric, and Bluenautilus:@Nyoome, ProtoplasmaKid, Gatoclass, Ouranista, and Deb:@Slowking4, Schyler, and Sasuke Sarutobi:. As all of us speak english, this idea could be tested on en.wp. Probably it would be useful to start with formerly very active users (>100 edits/month) that went inactive during the last year. We would need such a list of inactive users compiled by WMF. We would then need to establish an email text together, that would be sent to the users. Sending them manually by wikimail is possible, but if WMF can mass send the emails in the name of us endorsers, it would be less work. I guess the response rate will be <10 % (inactive users that answer to our email). We (endorsers) would need a non-public communication medium to (e.g. a mailing list, ...) to share the response emails and to analyse them collectively. I'm not sure what we need funding for, but we would need WMF's labour time in making the list and mass sending the emails. Maybe the other endorsers have better ideas on how to do this. Just my 2 cents, cheers, --Ghilt (talk) 20:22, 6 September 2018 (UTC)Reply
I agree with everything you say above. Deb (talk) 08:37, 7 September 2018 (UTC)Reply
We need to make sure items on the survey are reliable and valid, first. I think a random sample of all inactive users using a pilot survey would be better. Such a pilot survey might have around 50 questions. Then, after finding reliabilities (using Cronbach's alpha) and establish validity (using factor analysis), we could then send out the survey to other groups of inactive users. Schyler (talk) 19:33, 18 September 2018 (UTC)Reply
I agree with everything said so far. I must emphasize the need to protect the privacy of users who respond to our survey. — Mr. Guye (talk) (contribs)  01:52, 4 October 2018 (UTC)Reply
i would suggest qualitative survey methods, for example like Jackie Koerner [1] see also https://meta.wikimedia.org/wiki/Surveys -- Slowking4 (talk) 01:43, 31 October 2018 (UTC)Reply
Return to "IdeaLab/Email inactive (formerly active) users with a survey on inactivity reasons" page.