Background

The Android team has heard many requests from the community to ensure the edits made in the apps are of quality. Additionally, members of the community overall have asked for moderation tools that does not show performance preferences to a single language wiki, an example can be found in the 2022 wishlist. The team also needs to ensure we are not simply recreating an experience that exists, but enhancing workflows and addressing gaps in equity.

With the team's recent release of native watchlist, contribution history and edit history, as well as the addition of the undo and rollback button on the diff screen there is an opportunity to create a moderating solution in the app. The solution will leverage the strengths of workflows that exist in the app of microcontributions. We will also explore how to convert editors that are delivering value into patrollers. All of this will be done in partnership with the community, both existing administrators and prospective administrators. We will also work with the research team to improve ORES, and systems like it and the Moderator tools team who is tackling this problem space on Mobile Web, so that the solution isn't constrained to just the apps, rather that the apps is a complimentary experience and there is practical continuity for cross platform users.

Must Haves

• Editors are able to review diffs for a variety of edits
• Ability for editors to ‘revert’ or ‘skip’ suggestions
• Ability to display recent edits in varying wiki languages (doesn’t have to be same feed)
• Some form of filtering
• Allowing editors to create and store warning messages to accompany the revert
• A way to introduce the patrolling feature to experienced editors
• - Onboard experienced editors to patrolling on the app
• Instrumentation that helps us understand if the feature led to an increase in editors patrolling
• Instrumentation that helps us understand if the onboarding is sufficient
• Metric and Design goals that balances patrolling ease with importance of communication
• Metric and Design goals that balances clear vandalism and a good faith edit

Hypothesis & Assumptions

We believe that if there is an easier way for users to patrol tasks, and we can structure this in a way that it becomes an advanced-level task. We can give easier patrolling tasks (or a training task) to editors that are experienced, but do not have sysops or rollback rights and then uplevel them after reaching a threshold of initial patrolling activities. We believe to achieve this we must start with a proof of concept for general patrolling of edits based on users that have rollback rights. Finally, we assume that patrolling recent edits is desired as an initial priority and is more fitting for mobile than NPP, AFD, and AFC. Additionally, we assume there will be community consensus on a training task across languages.

How will we know we were successful

Validation

• Key Indicator 1: 65% of Target mature audiences that use the tool say they find it helpful for maintaining the quality of wikis and would recommend it to other patrollers (to be measured qualitatively, and then at 30-day point with survey)
• Key Indicator 2: 10% of target mature audiences interacted with another user by sending thanks or a talk page message.
• Key Indicator 3: 10% of target mature audiences engage with filter for preferences
• Key Indicator 4: 65% of Target mature audiences engage with the tool at least once in a thirty day window

Guardrails:

• Experienced users without rollback rights, users that have and have not used alternative patrolling tools equally understand the workflow (Completed with usability testing)
• We do not receive reports of tool being used to negatively target underrepresented content or contributors based on in-app reporting mechanisms

Curiosities:

• How does use of our tool compare to other patrolling tools when looking at MediaWiki Tags (SWViewer, Huggle, and Twinkle)
• Do we see an increase in Undo/Rollback/Thank events
• How popular is this task with our target audience relative to other Suggested Edits task?
• What actions are most popular in the feature?
• For Saved messages:
  • How often do users create messages from scratch instead of using an example message?
  • How often do users modify an example message before sending?
  • How often do users modify an example message and save that version to "Your Messages"?
  • How often do users click on each of the 10 example messages?
• For templates
  • How often are users using a Template while posting talk pages messages messages in Edit Patrol?
  • How often are they saving a message to "Your Messages" that contains a template?

User Stories

• As a Wikipedia Android app user that has rollback rights, I want to easily fight vandalism while on the train and save my progress, so that I can maintain the quality of wikis while on the go.
• As an accomplished experienced editor, I want to learn how to fight vandalism in a low stakes enviornment, so that I can earn admin privileges while still contributing in a way that enhances the quality of wikis, especially about topics that I am well informed about

Target Quant Regions and Languages

First Target Wiki

• Indonesian

Next Round Target Wikis

• French
• Chinese
• Spanish
• Igbo

Final Target before full scale

• English