Wikipedia:Bots/Requests for approval/HostBot 8

HostBot 8

Operator: (bot) Jtmorgan (talk · contribs · SUL · edit count · logs · page moves · block log · rights log · ANI search)

Operator (training modules): Ragesoss (talk · contribs · SUL · edit count · logs · page moves · block log · rights log · ANI search)

Time filed: 18:25, Tuesday, May 2, 2017 (UTC)

Automatic, Supervised, or Manual: Automatic (after supervised trial)

Programming language(s): Python

Source code available: https://github.com/jtmorgan/hostbot

Function overview: Posts a welcome message on new users' talk pages that includes links to introductory training modules on Programs & Events Dashboard, like this. Here's the template: {{Welcome training modules}}.

Links to relevant discussions (where appropriate): Wikipedia:Village_pump_(proposals)#Experiment_to_see_if_training_modules_are_helpful_for_new_users (permanent link). See also Wikipedia:Bots/Requests for approval/RagesossBot 3.

Edit period(s): Continuous

Estimated number of pages affected: 5000

Exclusion compliant (Yes/No): Yes

Already has a bot flag (Yes/No): Yes

Function details:

The main purpose of this task is to run a controlled experiment inviting new users to use the Programs & Events Dashboard training modules to learn the basics of Wikipedia. It finds recently registered accounts that have made between one and five edits, and sorts half of them into an experimental group and half into a control group. For the experimental group, it posts a welcome message (similar to what HostBot has done before with Teahouse invitations) that invites users to try the training modules, which are forked from the Wiki Ed classroom program training modules that we've been using and refining over the last few years and get very positive feedback from student editors.

Jonathan Morgan and I would like to get a sample of about 5000 invited users so that we can see if it makes a difference in terms of new users being more likely to stick around and keep editing. The last major experiment along these lines that I know of was The Wikipedia Adventure; in contrast to TWA, these trainings are more practical and wide-ranging, and have also been refined over time to try to head off the most common errors and confusing aspects of Wikipedia that new users run into.--ragesoss (talk) 18:25, 2 May 2017 (UTC)[reply]

The code has been tested on test.wikipedia.org and you can see a sample of the output here. J-Mo 18:32, 2 May 2017 (UTC)[reply]

Discussion

The discussion, archived here, has not been formally closed, but it seems that there is reasonable grounds for the trial. I certainly support such a trial, as editor retention is a key issue, that needs addressing. With new users, have you considered the best edit range? You said between 1 and 5, but I feel this is too low, and risks inviting vandalism only accounts into both groups. Perhaps a range of 7+ would filter these out? TheMagikCow (T) (C) 07:05, 5 May 2017 (UTC)[reply]