Page MenuHomePhabricator

Add instrumentation required to analyze engagement with talk page meta content (mobile)
Open, Needs TriagePublic

Description

In T312309, we are going to implement a link within with the DT-enhanced mobile talk page view.

When people tap this link, they will be shown the meta content (read: templates) that can sometimes appear within talk pages' lead sections.

This task involves the work with adding the instrumentation required to analyze how people are engaging with said link so that we can: A) decide whether it should be made more/less discoverable and B) understand what – if any – impact people tapping this link has on the change(s) they go on to subsequently publish to the talk page on which the link appears.

Research questions

These are the questions the instrumentation we add will need to help us answer.

  1. On mobile talk pages that contain meta content (read: the "About this page" link is shown), how does the quality of talk page edits vary between people who view this meta content and people do not view it?
  2. Of all the visits people make to mobile talk pages that contain the "About this page" link, what percentage of people tap this link? How does engagement (read: taps) with the "About this page" link compare to engagement with other affordances on the page?

Instrumentation requirements

⚠️TBD⚠️

Done

  • ===Instrumentation requirements are documented that will enable us to answer the ===Research questions
  • ===Instrumentation requirements are implemented

Event Timeline

ppelberg moved this task from Untriaged to Upcoming on the Editing-team board.
ppelberg moved this task from Backlog to Triaged on the DiscussionTools board.
ppelberg added subscribers: MNeisler, DLynch.
  1. On mobile talk pages that contain meta content (read: the "About this page" link is shown), how does the quality of talk page edits vary between people who view this meta content and people do not view it?

You should be aware that some tmboxes (=the content shown in MobileFrontend’s “About this page” overlay) should affect talk page edits’ quality more or less directly (e.g. talk page headers, maybe archive boxes as well—if some topic has already been discussed, restarting the discussion may get reverted), but others like WikiProject boxes or VRT permission templates shouldn’t (they are metadata of the article, not instructions for users of the talk page). Maybe the instrumentation should record not only the presence or absence of the overlay, but also types of templates included in it.