An R package for recovering and estimating subjective beliefs from Quadratic Scoring Rule (QSR) belief elicitation tasks, accounting for risk preferences. Implements the theory of Harrison & Ulm (2016) and the behavioral estimation approach of Andersen, Fountain, Harrison, and Rutström.
devtools::install_github("bamonroe/rcbelief")- Belief recovery — Recover subjective beliefs from QSR reports given estimated risk preferences (
rec_bel) - Interval regression — Jointly estimate recovered beliefs and risk preferences through an interval regression approach
- Behavioral estimation — Jointly estimate subjective beliefs and risk preferences using a structural behavioral model
- CRRA:
x^(1-r) / (1-r) - Power:
x^r
- EUT (expected utility, no weighting)
- Power (Quiggin 1982)
- Inverse-S
- Two-parameter Prelec (1998)
MIT