Perceived correlations in risk attitudes

Last registered on June 26, 2022

View Trial History

Pre-Trial

Trial Information

General Information

Title

Perceived correlations in risk attitudes

RCT ID

AEARCTR-0009404

Initial registration date

June 22, 2022

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

June 26, 2022, 5:23 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country

Singapore

Region

Singapore

Primary Investigator

Name

Jingcheng Fu

Affiliation

National University of Singapore

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Songfa Zhong

PI Affiliation

National University of Singapore

Contact Investigator

PI Name

Wencong Li

PI Affiliation

National University of Singapore

Contact Investigator

Additional Trial Information

Status

In development

Start date

2022-06-27

End date

2022-08-26

Keywords

Behavior

Additional Keywords

belief, noise, correlation, risk preferences, experiment

JEL code(s)

C91, D81, D83, D84

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

We investigate the perceived links between risk preferences within and across domains defined over probability (low/moderate) and outcome (prospect/hazard). In an experimental setting, we first elicit subjects' risk taking behaviors in different domains, and then measure their beliefs about links of behaviors within the same domain and across different domains. We test whether subjects overestimate within-domain links (i.e., underestimate noise), and document how they perceive the cross-domain links. We further examine whether subjects give sufficient decision weights to the corresponding within-domain links when they evaluate a cross-domain link.

External Link(s)

Registration Citation

Citation

Fu, Jingcheng, Wencong Li and Songfa Zhong. 2022. "Perceived correlations in risk attitudes." AEA RCT Registry. June 26. https://doi.org/10.1257/rct.9404-1.0

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

The experiment consists of two parts. The first part measures risk preferences in four domains using the standard choice list whereby subjects make decisions between the lottery and a list of sure amounts. In the second part, which is the main focus of the experiment, subjects provide incentivized judgments of the links between lotteries within the same domain and across different domains. The links are presented in the form of how well the choices for lottery X predict those for lottery Y.

Intervention Start Date

2022-06-27

Intervention End Date

2022-08-26

Primary Outcomes

Primary Outcomes (end points)

The perceived within-domain PCs (two guesses for each of the four domains per subject);
The perceived cross-domain PCs (two guesses for each of the six pairs of domains per subject).

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

Risk preferences elicited using choice lists; questionnaire responses (confidence, risk preferences, demographics and academic background).

Secondary Outcomes (explanation)

Experimental Design

The experiment consists of two parts. In Part 1, we elicit the certainty equivalents (CEs) for binary lotteries in four domains based on probability and outcome: moderate prospect, long-shot prospect, moderate hazard, and long-shot hazard. The CEs are elicited using standard multiple price lists. 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. The Part 1 decisions are converted into ratings for the lotteries. For each lottery, we count the number of times a subject chooses the lottery over sure amounts as their rating. A subject with a higher rating is more willing to take risk for this lottery, compared to one with a lower rating.

In Part 2, subjects are asked to express their perceptions of the links between ratings of different lotteries using the percent concordant (PC). This number answers the question that, suppose a pair of subjects are taken at random, what is the probability that the one who has a higher rating for one lottery also has a higher rating for another lottery. With four domains of lotteries, there are four within-domain links and six cross-domain links. We ask subjects to predict each link twice. One of the 20 questions is randomly selected to determine whether subjects receive S$15 using a Binarized Scoring Rule.

Experimental Design Details

Randomization Method

Randomization of question format and order is implemented by Qualtrics.

Randomization Unit

Individual

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

300 subjects.

Sample size: planned number of observations

2,400 (300 subjects × 4 domains × 2 duplicates) guesses of within-domain links; 3,600 (300 subjects × 6 domain pairs × 2 duplicates) guesses of cross-domain links.

Sample size (or number of clusters) by treatment arms

300 subjects (one treatment).

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

We are able to detect a significant difference with significance level of 0.05 and power of 0.8 when the guesses deviates from the true value by more than 0.16 SD using one-sample t-test.

Supporting Documents and Materials

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

IRB