Verbal descriptions of risks in financial contexts

How to cite

Becker, Christoph; Trautmann, Stefan T.; Nolte, Julia. 2026. "Verbal descriptions of risks in financial contexts." AEA RCT Registry. AEARCTR-0019378. https://doi.org/10.1257/rct.19378-1.0

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

Christoph Becker* , Stefan T. Trautmann , Julia Nolte

* Primary investigator

Registry ID
AEARCTR-0019378
DOI
https://doi.org/10.1257/rct.19378-1.0
First registered
Last updated
Status
In development
Trial dates
2026-08-17 – 2026-08-31
Countries
Private
JEL codes
C83, D81, D84
Registry record
http://www.socialscienceregistry.org/trials/19378

Abstract

Risks are often communicated in financial contexts using verbal categories (e.g. “very conservative” or “very risky”) for different ranges of risk. Previous studies have explored how decision makers interpret these labels, finding a large amount of heterogeneity in risk perceptions. In this project we study the reverse – how do decision makers themselves verbally communicate numerical risks. We draw data from an exploratory survey in which respondents are given 6 different numerical risks and are asked to describe them without using numbers or numerical terms. These 6 risks range from relatively small (2%) to quite large probabilities (95%). Respondents are presented these risks in either an ascending, descending or random order. We are primarily interested which terms and descriptions are commonly used for a given risk. We further link this to demographic factors, as well as financial participation and risk attitude.

Intervention

Eliciting descriptions of numerical risks using open text boxes.

Experimental design

Participants in the survey are presented with a fictional scenario in which they are asked to consider the risk of their retirement savings getting depleted prematurely. In 6 different questions based on this scenario, they are asked to describe specific percentage risks (e.g. 2%) in their own words.

Participants face these 6 risk questions in 3 different treatment orders, which vary of the risks are presented randomly or in ascending or descending order.

After the main task, participants answer a set of additional questions, as well as basic demographics (age, gender, education, and income). Participants are recruited from a representative US sample (in terms of age, gender, race and political orientation) and earn a flat payment for their participation.

Design details

Randomization method
Three treatments, treatments are alternated between participants, participants enter study independently.
Randomization unit
Individual
Observations
600 Prolific participants
Arms
200 participants per treatment
Primary outcomes
(i) Descriptions of numerical risks (string).
Secondary outcomes
(i) Financial participation (ii) Confidence (iii) Risk tolerance (iv) Financial literacy (v) Demographics (age, gender, education, income)

Cited by

No citation data available for this registration yet.

Keywords