“Banner Ads Considered Harmful § EVSI”, Gwern2017-01-08 (R, ads, Bayes, decision theory, power analysis; similar)⁠:

9 months of daily A/B-testing of Google AdSense banner ads on Gwern.net indicates banner ads decrease total traffic substantially, possibly due to spillover effects in reader engagement and resharing.

Demo code of simple Expected Value of Sample Information (EVSI) in a JAGS log-Poisson model of traffic (which turns out to be inferior to a normal distribution for 2017 traffic data but I keep here for historical purposes).

• Modeling Effects of Advertising: Global rather than Local
• Implementation: In-Browser Randomization of Banner Ads
• Ads As Decision Problem
  • Ad Harms
    • Replication
      • Pandora
      • Mozilla
      • LinkedIn
      • McCoy Et Al 2007
      • Google
      • PageFair
      • Yan Et Al 2020
    • Aral & Dhillon2020
    • Suárez & García-Mariñoso2021
    • They Just Don’t Know?
• Design
  • Descriptive
  • Power Analysis
    • NHST
    • Bayesian
• Analysis
  • Descriptive
  • Simple Tests & Regressions
  • Stan ARIMA Time-Series Model
  • Decision
• Discussion
• Followup Test
  • Design
    • Implementation
  • Analysis
• External Links
• Appendices
  • Stan Issues
  • Stan: Mixture Time-Series
  • EVSI

Similar Links: