THanks @YLiou_WMF for creating this task. I co-assigned @CMyrick-WMF as she'll do a large portion of the analysis.

Weekly update:

• Queried the QS data
• Wrangled response data for analysis next week: cleaned; accounted for dupes, click-backs; reshaped
• Next week: join with session data; begin stage ii analysis

Weekly update:

• Preliminary survey response results shared with T393802 research team.
• Queried and joined session data with survey responses
• Queried and joined topics data
• Began Stage II Analyses

Weekly update:

• Survey data has been weighted by OS family, browser family, referrer type, geography (by country), browser language preference, whether user was reading an article in the top 0.05% of enwiki articles by traffic (to adjust for "spiky" traffic)
• Session length and article topic data analysis begun
• Phase II survey analysis incorporated into draft readers foundational report deck

Weekly update:

• Debugging
• Determined thresholds and other rules for sessions
  • max 20 pages per session (99% percentile for random sample session length is 20 pages)
  • max 28800 seconds per session (8 hours; since avg online time is 7 hours worldwide and 7+ hours in U.S.)
  • exclude immediately-sequential pageview of prior page_id if same referrer class
  • if multiple access methods, browsers, or devices logged, take MIN timestamp-wise
• Queried pageview data for top 0.05% top-popular-pages per day; used to code pageviews as a view of a top-popular page or not
• Finished coding rules related to language preference
• Began compiling control data based on random sample of pageview traffic
• Began updating the weighting code
• QCing in progress

Weekly update:

• Finished QCing
• Finished compiled control data based on random sample of pageview traffic
• Calculated and applied weights to the respondent data, using the control data for the targets
• Finished documenting all thresholds, rules, logic, etc applied during the data cleaning

Weekly update:

• survey data analysis is continuing to be added to deck for stakeholders
• preliminary presentations of survey data conducted for stakeholders this week
• final deck expected to be shared next week (and hypothesis for T393802 closed)

Weekly update:

• Queried enwiki QS data
• Wrangled response data for analysis next week
• Next week: join with session data; begin stage iii/iv analyses

Weekly update:

• For Global Readers Survey English
  • Joined with session data
  • Completed state iii/iv analyses
  • Completed calculation and application of weights
  • Currently investigating possible data loss (between QS and LS)

• For Global Readers Survey non-English (T410918)
  • Queried QS data for subset
  • Wrangled response data for subset
  • Pulled control data (for weighting) for subset
  • Next week: complete queries and wrangling for all surveyed wikis; join with session data; begin stage iii/iv analyses