Loss aversion: loss looms larger or gain looms smaller?

How to cite

LI, Wencong; ZHONG, Songfa. 2023. "Loss aversion: loss looms larger or gain looms smaller?." AEA RCT Registry. AEARCTR-0010861. https://doi.org/10.1257/rct.10861-1.0

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

Wencong LI* , Songfa ZHONG

* Primary investigator

Registry ID
AEARCTR-0010861
DOI
https://doi.org/10.1257/rct.10861-1.0
First registered
Last updated
Status
In development
Trial dates
2023-03-08 – 2023-05-31
Countries
Singapore (Singapore)
JEL codes
C91, D81
Registry record
http://www.socialscienceregistry.org/trials/10861

Abstract

Loss aversion is a critical contribution of Prospect Theory to the study of decision-making. However, previous research on the construct of loss aversion in decisions under risk unveils mixed results. Some studies provide evidence for loss aversion, whereas others show a tendency for gains to be the same impactful as losses. We will employ two representations, gain-first framed and loss-first framed questions, to explore this discrepancy. In an experimental setting, we will elicit subjects’ general attitudes towards risks and losses through binary-choice questions. We will further test whether the lack of evidence supporting loss aversion in some studies is attributed to outcomes framed as gain exhibited first in mixed lotteries.

Intervention

The experiment involves two treatment conditions. The first treatment condition is that the relatively lower outcome is exhibited first for binary-outcome lotteries; and the second treatment condition is that relatively higher outcome is exhibited first for binary-outcome lotteries.

Additionally, the experiment consists of two sets of binary choices. The first set of binary choices measures risk preferences on the domain of gains and on the domain of losses separatly. The second set of binary choices, which is the main focus of the experiment, measures subjects’ sensitivities to losses compared to gains using mixed prospects.

Experimental design

The experiment involves two treatment conditions. The first treatment condition is that the relatively lower outcome is exhibited first for binary-outcome lotteries; and the second treatment condition is that relatively higher outcome is exhibited first for binary-outcome lotteries. Each subject will be randomly allocated to one of the treatment groups and finish a set of binary-choice questions and a post-experimental questionnaire.

We use the similar choice paradigm ascribed by Kahneman & Tversky, (1979). The set of binary-choice questions involve two-outcome prospects in three domains based on outcome: positive prospects, negative prospects and mixed prospects. Subjects will indicate their preference for two alternatives in each question. Aggregating over all questions, we count the choices in the direction predicted by fourfold pattern and loss aversion in each treatment group.

Losses are deducted from the participation fee that subjects receive at the beginning of the experiment to ensure that the overall payment from the experiment is never negative.

After subjects complete the set of binary-choice questions, they will fill in a questionnaire that collects data on the following characteristics:
1. Confidence for decisions
2. Cognitive reflection test;
3. General risk attitude;
4. Demographics questions.

(REFERENCE: Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 1, 143–172. https://doi.org/10.2307/1914185)

Design details

Randomization method
Randomization of question format and order is implemented by Qualtrics.
Randomization unit
Individual.
Clusters
250 subjects.
Observations
2250 (125 subjects x 18 questions) decisions in each treatment group.
Arms
125 subjects/group (two treatment groups).
Primary outcomes
The estimates of the loss-aversion parameter under the "lower-outcome-first" treatment condition; The estimates of the loss-aversion parameter under "higher-outcome-first" treatment condition.
Secondary outcomes
Risk preferences measured on the domain of gains and on the domain of losses; questionnaire responses (confidence, risk preferences, demographics background).

Cited by

No citation data available for this registration yet.

Keywords