Revealing a Reference Dependent Preference for Mixing: An Experimental Study on Risk Preferences over Gains and Losses

How to cite

Feldman, Paul; Rehbeck, John. 2019. "Revealing a Reference Dependent Preference for Mixing: An Experimental Study on Risk Preferences over Gains and Losses." AEA RCT Registry. AEARCTR-0004214. https://doi.org/10.1257/rct.4214-1.0

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

Paul Feldman* , John Rehbeck

* Primary investigator

Registry ID
AEARCTR-0004214
DOI
https://doi.org/10.1257/rct.4214-1.0
First registered
Last updated
Status
On going
Trial dates
2019-05-13 – 2020-06-30
Countries
Private
JEL codes
C91, D81, D91
Registry record
http://www.socialscienceregistry.org/trials/4214

Abstract

We explore the effect of two distinct frames on preferences for mixtures. Given the pervasive evidence of a preference for mixtures---non-degenerate distributions over two lotteries---we study the effect of two different frames on this behavior. The distinctive feature of these frames is whether outcomes are presented as losses or gains, while final wealth outcomes remain the same under both treatments. We expect an increase in loss-loving behavior, which would be consistent with the prior findings.

Experimental design

This is a lab experiment in which we elicit preferences over lotteries. Lotteries are limited to a linear convex budget in the probability simplex over three outcomes $2, $10, and $30. For the main treatment, we endow subjects with $30 and instead 'frame' outcomes as -$28, -$20, and $0. Therefore, both treatments have identical outcomes, but they are presented as either losses or gains. We have two treatment groups, the first is exposed to the loss-frame and then the gain-frame. For our other group, we reverse the order of the frames.

Design details

Randomization method
Random task order. 2 treatment groups. Pay on a decision at random. All risk is resolved using a ten-sided die.
Randomization unit
Between-subjects and Within-Subjects.
Clusters
2
Observations
164-192
Arms
82 in each treatment for between subject, 164 within-subject analysis.
Primary outcomes
Changes in mixing behavior across frames.

Cited by

No citation data available for this registration yet.

Keywords