The Large Number of Incompatible Choices Implied by Representing Risk Preference Through Curvature

How to cite

Martinez, Seung-Keun; Augenblick, Ned. 2023. "The Large Number of Incompatible Choices Implied by Representing Risk Preference Through Curvature." AEA RCT Registry. AEARCTR-0012180. https://doi.org/10.1257/rct.12180-1.1

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Seung-Keun Martinez* , Ned Augenblick

* Primary investigator

Registry ID
AEARCTR-0012180
DOI
https://doi.org/10.1257/rct.12180-1.1
First registered
Last updated
Status
Completed
Trial dates
2023-10-11 – 2023-10-17
Countries
United States of America ()
JEL codes
D90, C90, D81
Registry record
http://www.socialscienceregistry.org/trials/12180

Abstract

In preceding theoretical work, we develop methodology to measure the implied (expected) utility curvature from choices over objective lotteries. In this incentivized survey, we study the consistency of implied utility curvature over simple objective lotteries.

Experimental design

Each participant will be required to make 10 2-part choices over uncertainty (described above in primary outcomes). The Lotteries are follows.

Lottery A: 1.00*$20
Lottery B: 0.50*$10+0.50*$30

Lottery A: 1.00*$40
Lottery B: 0.50*$30+0.50*$50

Lottery A: 1.00*$12
Lottery B: 0.90*$10+0.10*$30

Lottery A: 1.00*$32
Lottery B: 0.90*$30+0.10*$50

Lottery A: 1.00*$28
Lottery B: 0.10*$10+0.90*$30

Lottery A: 1.00*$48
Lottery B: 0.10*$30+0.90*$50

Lottery A: 0.75*$20+0.25*$40
Lottery B: 0.25*$10+0.75*$30

Lottery A: 0.75*$40+0.25*$60
Lottery B: 0.25*$30+0.75*$50

Lottery A: 0.50*$20+0.50*$40
Lottery B: 0.125*$10+0.75*$30+0.125*$50

Lottery A: 0.50*$40+0.50*$60
Lottery B: 0.125*$30+0.75*$50+0.125*$70

We will also collect basic demographic information via prolific, and we will ask people to state their agreement to the the following statements on a likert-scale:

I felt certain about my lottery evaluations.
I am comfortable with financial risk.
I am better at math than most people.

Design details

Randomization method
Automated randomization by Qualtrics.
Randomization unit
Within-subject randomization at the individual level.
Clusters
500 individual participants. Since each individual will answer 10 risk elicitation questions, we will treat each participant as a "cluster."
Observations
10 observations per participant for a total of 500 observations.
Arms
Within subject randomization of risk elicitation question. All participants will see all risk elicitation questions.
Primary outcomes
We will be eliciting preferences over objective lotteries. The elicitation device consists of two steps. 1) A binary choice between Lottery A and Lottery B. 2) A choice of how much guaranteed money must be taken away from the preferred lottery to make the decision maker indifferent between the two lotteries. The choice described above will be incentivized through the standard Becker-DeGroot-Marschak mechanism.

Cited by

No citation data available for this registration yet.

Keywords