Regulating Risk: A Stress-Related Intervention for Loss Aversion

How to cite

Grebitus, Carola; Cetik, Nilufer. 2026. "Regulating Risk: A Stress-Related Intervention for Loss Aversion." AEA RCT Registry. AEARCTR-0017757. https://doi.org/10.1257/rct.17757-1.0

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

Carola Grebitus* , Nilufer Cetik

* Primary investigator

Registry ID
AEARCTR-0017757
DOI
https://doi.org/10.1257/rct.17757-1.0
First registered
Last updated
Status
In development
Trial dates
2026-01-26 – 2026-02-13
Countries
Private
JEL codes
C91, D81, D91
Secondary IDs
OSF: https://osf.io/ye9k2
Registry record
http://www.socialscienceregistry.org/trials/17757

Abstract

This study examines whether a brief Non-Sleep Deep Rest (NSDR) session can reduce loss aversion, a common cognitive bias where people tend to react more strongly to potential losses than to equivalent gains. Loss aversion plays a major role in behaviors related to finances, health, and risk-taking, and it is often heightened under stress.
We will conduct a laboratory experiment with approximately 600 adult participants. Each participant will be randomly assigned to one of two groups. The treatment group will listen to a short NSDR audio recording, while the control group will listen to a neutral audio recording of equal length. Before and after the audio session, participants will complete a series of economic choices involving real monetary gains and losses. These choices will be used to estimate a loss aversion coefficient for each participant. We will then compare changes in behavior between the two groups. Self-reported stress levels will also be measured before and after the intervention to examine whether reductions in stress explain any change in decision-making.
In a secondary task, participants will select a snack from a small, medium, or large set of options. After the audio session, they will be reminded of their original selection and given the opportunity to either confirm or change it. We will examine whether the complexity of the choice set and the NSDR session influence switching behavior, especially with regard to changes between healthy and unhealthy options.
The goal of this study is to evaluate whether a low-cost, non-pharmacological intervention like NSDR can improve economic decision-making and support self-regulation under conditions of stress.

Intervention

Participants in the treatment group will experience a guided audio session based on Non-Sleep Deep Rest (NSDR), a form of deep relaxation that involves being still with eyes closed while listening to calming, structured instructions designed to reduce stress and increase focus. The NSDR session lasts approximately 10 minutes and is delivered via headphones in a quiet laboratory setting. Participants in the control group will listen to a neutral audio recording of the same duration. This recording contains emotionally neutral content to match the treatment condition in terms of length, tone, and attention demands, but without the relaxation or stress-reducing elements of NSDR.
Both audio recordings are presented before participants continue to complete a series of economic and behavioral decision-making tasks. The intervention is designed to test whether brief exposure to NSDR influences how individuals evaluate risk and regulate their choices under stress.

Experimental design

Participants will be randomly assigned to one of two audio conditions and will complete a series of decision-making tasks involving monetary choices. Participants will also complete a behavioral choice task involving snack selection. The study will take place in a controlled laboratory environment, and group assignment will be randomized using a computer-based procedure. All procedures have been approved through the appropriate ethics review process.

Design details

Randomization method
This is a randomized controlled lab experiment designed to test whether a brief Non-Sleep Deep Rest (NSDR) audio intervention influences loss aversion, probability weighting, and decision stability under stress. Participants are randomly assigned by computer to one of two groups: the NSDR treatment or a neutral audio control. Randomization is stratified based on pre-intervention stress levels, assessed through a short questionnaire at the beginning of the session. Participants also take part in a snack choice task. They are randomly assigned to one of three choice set sizes (4, 9, or 16 snack options) by the computer, make an initial selection before the audio intervention, and are given a chance to revise or confirm their choice afterward.
Randomization unit
The unit of randomization is the individual participant. Each participant is independently assigned by computer to either the NSDR treatment group or the neutral audio control group. Randomization is stratified based on baseline stress level, measured at the start of the session. Additional randomization occurs at the individual level for two secondary components: Snack choice set size (4, 9, or 16 options) and incentivization of the Oprea (2024) lottery task, where one in ten participants is randomly selected for real payout.
Clusters
600 individuals will be randomly assigned into control and treatment groups. No clusters are defined.
Observations
600 individuals
Arms
1) 300 individuals in control group (neutral audio recording), 300 individuals in treatment group (NSDR audio recording) 2) 200 individuals will have 4 options, 200 individuals will have 9 options, 200 individuals will have 16 options available for snack choices
Primary outcomes
The main outcome of interest is the loss aversion coefficient estimated from participants’ choices in the lottery task. This captures how strongly individuals weigh potential losses relative to gains when making decisions under risk. A second key outcome is the probability weighting parameter, which reflects how participants perceive and respond to different levels of probability in risky decisions. Both parameters will be estimated using a structural model based on Prospect Theory. We will also measure within-subject change in loss aversion from pre- to post-intervention, and compare these changes between the treatment and control groups. This allows us to test whether the intervention reduces loss aversion relative to baseline.
Secondary outcomes
As a secondary behavioral outcome, we will analyze snack choice switching, specifically whether participants switch from a healthy to an unhealthy option (or vice versa) after the intervention. This helps us explore the broader impact of the intervention on decision stability and self-control.

Cited by

No citation data available for this registration yet.

Keywords