Detailed information can be found in the pre-analysis plan.
Experimental design
See pre-analysis plan.
Design details
Randomization method
By computer.
Randomization unit
The randomization unit is the HTTP-Cookie of the website visitor.
Clusters
The expected number of website visitors is unknown. The sample size for the analysis on the intensive margin depends on the treatment effects on the extensive margin. If the treatments do not affect the extensive margin, I expect to observe 300,000 transactions. For more details, see pre-analysis plan.
Observations
The expected number of website visitors is unknown. The sample size for the analysis on the intensive margin depends on the treatment effects on the extensive margin. If the treatments do not affect the extensive margin, I expect to observe 300,000 transactions. For more details, see pre-analysis plan.
Arms
If the treatments do not affect the extensive margin, I expect to observe 30,000 transactions per treatment group for the main analysis on the intensive margin. For more details, see pre-analysis plan.
Primary outcomes
- Probability to purchase at the shop. (extensive margin)
- Probability to buy the carbon offset conditional on purchasing at the shop. (intensive margin)
- Answers to two survey questions (on a 7-point Likert scale).
Secondary outcomes
- Number of follow-up purchases.
- Number of beverage crates ordered.
Cited by
No citation data available for this registration yet.