Progress Update (13-June)

• Attended meeting with Moderator Tools and Research teams, where we discussed the current status of the patrolling dataset and dashboard, and the next steps needed to measure moderation actions and finalize the KR
• Met with Pablo on the Research team to get a more detailed overview of the current patrolling dataset and learn more about what can be measured with the current data and any limitations. I confirmed the dataset was in the format needed for analysis.
• Began to explore the patrolling dataset and ran initial queries to collect data on moderation trends based on the latest snapshot available, including the number of patrolling actions completed a month.

Progress Update (27-June)

• Continued to explore the patrolling dataset and review the preliminary report on Wikipedia Patrolling measurement.
  • Provided comments and questions on the report
  • Continued to identify and summarize potential insights that can be used to help inform the KRs.
• Began to identify a possible target for the KR[1]. As we’ve adjusted the KR from looking at overall increases in moderation actions to more specifically evaluating the use of the homepage, I’ve begun to review data for comparable moderator activities and targets achieved by similar past experiments.

[1] New proposed KR: "By the end of Q3, X% of contributors who had an engaged session on the homepage returned to have another engaged session on the homepage in their second week."

Progress Update (1-July)

• We identified possible targets for KR based on an exploration of moderator trends using the new patrolling dataset created by Research in T392210 and insights from past interventions and experiments.
• A KR target of 15% was proposed based on the second week retention rate of patrollers using the FlaggedRevs extension at English Wikipedia. While not an exact baseline for moderator homepage retention, this was recommended as a comparable moderation retention rate metric.
• Additional findings into moderator retention rates, the number of moderation actions, and other moderator trends were summarized in a document. This document can be referenced for future discussions on defining an engaged session and refining KR targets.

Next steps
We need to define what actions will be included in the definition of “engaged session” on the homepage to be offered to moderators before finalizing the KR. Work to define these actions will be supported by findings from the new Wikipedia patrolling dataset and completed under a new hypothesis.

Code Repo