Examine the [[https://meta.wikimedia.org/wiki/Schema:MobileWebSectionUsage | data]] from the experiment set up in T120292 , where 0.05% of mobile web users were shown all pages with every section expanded on initial load, alongside a control group of 0.05% that kept seeing the standard view where all sections all initially collapsed. (For the most part this is not an A/B test in a strict sense, because we are not comparing the same metric for both groups.) This is a followup to T118041, which preceded that experiment.
The goal is to inform the discussion about the tradeoffs of collapsing sections by default on mobile.
[] Compare (in a suitable sense) how many sections are scrolled into view in the uncollapsed view with how many are opened in the collapsed view
[] same when restricted to scroll/open events that are not followed by another event within the next 5 seconds (to distinguish focused reading from navigation activities)
[] Take into account what we know about usage of separate TOC on mobile where available (e.g. from Android instrumentations for [[https://meta.wikimedia.org/wiki/Schema:MobileWikiAppToCInteraction |TOC]] and [[https://meta.wikimedia.org/wiki/Schema:MobileWikiAppPageScroll |scroll]] usage?)
[] ~~Compare user session lengths (defined as the number of pages viewed)?~~ (not possible due to apperent bug regarding the sessionId field, see T128931#2283390)
[] Compare pageview session durations (defined e.g. as the time between "entered" and the last event logged on the same page)?