Heuristics and the ErgEx experiment
This project explores to what extend simple heuristics could have produced the choices made by participants in the controlled ErgEx decision environments. Which simple heuristics are approximately growth rate optimal and which best explain observed participant choices?
Why It Matters
Controlled experiments make it possible to compare behavioral data with specific heuristic models. If different dynamics elicit different heuristics, that would support the view that people adapt their decision rules to environmental structure.
Possible Starting Points
- Compile participant-level choice data from CPH experiments.
- Specify candidate heuristics.
- Fit or classify participants by heuristic.
- Compare heuristic prevalence across environments.
- Project lead
- Colm Connaughton, Emilie Soysal
- Collaborators
- Ratin Sazedul
Deliverables
- Data analysis note
- Behavioral modeling paper
- Reusable heuristic-fitting code
Created 2025