Imagining the Future: Experimental Evidence

How to cite

Coffman, Katherine; Bordalo, Pedro; Gennaioli, Nicola; Burro, Giovanni; Shleifer, Andrei. 2023. "Imagining the Future: Experimental Evidence." AEA RCT Registry. AEARCTR-0011344. https://doi.org/10.1257/rct.11344-1.0

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

Katherine Coffman* , Pedro Bordalo , Nicola Gennaioli , Giovanni Burro , Andrei Shleifer

* Primary investigator

Registry ID
AEARCTR-0011344
DOI
https://doi.org/10.1257/rct.11344-1.0
First registered
Last updated
Status
In development
Trial dates
2023-05-15 – 2023-05-17
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/11344

Abstract

This project aims to understand how people form beliefs about future events, particularly unfamiliar or novel events. We propose that memories of past experiences, both relevant and potentially irrelevant, shape these beliefs. Suppose an individual is estimating the likelihood of a significant cyberattack in the future. We argue that the individual will rely, in part, on past experiences that have some similarity to the event they are trying to imagine (a significant cyberattack). Our theory proposes that individuals who have personal experience with very similar things – for instance, identity theft – will have an easier time imagining a future cyberattack, boosting their estimates of the likelihood of it occurring. We will run a controlled experiment that investigates these memory mechanisms, exploring how exogenously varying which past (similar but objectively irrelevant) experiences “come to mind” impacts beliefs about the likelihood of future events.

Intervention

Participants will be randomly-assigned to one of four treatments:
• ID Theft prime
• Personal financial struggle prime
• Loss of a loved one prime
• Control treatment

For individuals in the ID theft prime, participants will be asked whether they have personal experience with identity theft. If they do, they will answer a few short questions about the experience, prompting them to remember it. Then, they will write a few short sentences about what they recalled when thinking about the experience.

For individuals in the financial struggles prime, participants will be asked whether they have personal experience with financial difficulties. If they do, they will answer a few short questions about the experience, prompting them to remember it. Then, they will write a few short sentences about what they recalled when thinking about the experience.

For individuals in the loss of loved one prime, participants will be asked whether they have lost a loved one. If they do, they will answer a few short questions about the experience, prompting them to remember it. Then, they will write a few short sentences about what they recalled when thinking about the experience.

Individuals in the Control treatment do not see an experience prompt of this form.

After this part of the survey, we provide participants with some basic information about cyberattacks. Then, we ask them to estimate the likelihood of a significant cyberattack impacting them or people like them in the next five years in the United States. We then ask participants to describe what they thought about when imagining the possibility of a cyberattack and how vividly they imagined it.

Then, we ask all participants about whether they’ve had each of a series of negative life experiences (just yes/no), including the three possible primed experiences, and how similar they feel these experiences are to a cyberattack. Finally, participants answer demographic questions.

Experimental design

We will recruit participants on Prolific. They will be asked to complete a 15 minute survey.

Participants will be randomly-assigned to one of four treatments:
• ID Theft prime
• Personal financial struggle prime
• Loss of a loved one prime
• Control treatment

For individuals in the ID theft prime, participants will be asked whether they have personal experience with identity theft. If they do, they will answer a few short questions about the experience, prompting them to remember it. Then, they will write a few short sentences about what they recalled when thinking about the experience.

For individuals in the financial struggles prime, participants will be asked whether they have personal experience with financial difficulties. If they do, they will answer a few short questions about the experience, prompting them to remember it. Then, they will write a few short sentences about what they recalled when thinking about the experience.

For individuals in the loss of loved one prime, participants will be asked whether they have lost a loved one. If they do, they will answer a few short questions about the experience, prompting them to remember it. Then, they will write a few short sentences about what they recalled when thinking about the experience.

Individuals in the Control treatment do not see an experience prompt of this form.

After this part of the survey, we provide participants with some basic information about cyberattacks. Then, we ask them to estimate the likelihood of a significant cyberattack impacting them or people like them in the next five years in the United States. We then ask participants to describe what they thought about when imagining the possibility of a cyberattack and how vividly they imagined it.

Then, we ask all participants about whether they’ve had each of a series of negative life experiences (just yes/no), including the three possible primed experiences, and how similar they feel these experiences are to a cyberattack. Finally, participants answer demographic questions.

Design details

Randomization method
Randomization by Qualtrics program
Randomization unit
Individual
Clusters
3000 individuals
Observations
3000 individuals
Arms
3000 individuals spread evenly across the 4 treatments, approximately 750 per treatment
Primary outcomes
Our key outcome variable is the estimated likelihood of a significant cyberattack. We have two of these likelihood questions: (i) What do you believe is the likelihood that you will be significantly impacted by a cyberattack over the next 5 years? Please indicate on a scale of 0 - 100, where 0 indicates that there is no chance at all that you will be significantly impacted by a cyberattack and 100 indicates that you will definitely be significantly impacted. (ii) Now, think of 1,000 people just like you in the United States. Out of those 1,000 people, how many do you believe will be significantly impacted by a cyberattack over the next 5 years? We will analyze both outcome measures independently. In addition, we will construct an index outcome for each participant by constructing a z-score for each of the two individual outcome questions and averaging them.
Secondary outcomes
We are also interested in the self-reported vividness with which they imagined a cyberattack (reported on a 1 - 7 scale).

Cited by

No citation data available for this registration yet.

Keywords