Lies, Truths, and Perceptions

Last registered on February 05, 2025
View Trial History

Pre-Trial

Trial Information

General Information

Title
Lies, Truths, and Perceptions
RCT ID
AEARCTR-0015324
Initial registration date
February 03, 2025

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
February 05, 2025, 9:17 AM EST

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

Locations

Country
United States of America
Region

Primary Investigator

Name
Benjamin Ho
Affiliation
Vassar College
Contact Primary Investigator

Other Primary Investigator(s)

PI Name
Olga Shurchkov
PI Affiliation
Wellesley College
Contact Investigator
PI Name
Tatyana Deryugina
PI Affiliation
University of Illinois at Urbana-Champaign
Contact Investigator

Additional Trial Information

Status
In development
Start date
2025-02-05
End date
2026-02-25
Keywords
Behavior
Additional Keywords
Disinformation, Deception, Lie Detection, Political Communication
JEL code(s)
C91, D91, D72
Secondary IDs
https://osf.io/bj3d9
Prior work
This trial is based on or builds upon one or more prior RCTs.
Abstract
Disinformation is a growing concern in today's digital environment, influencing public opinion on social media, political discourse, and everyday decision-making. This study investigates individuals’ ability to discern the truthfulness of self-reports made by anonymous participants, focusing on how the extremeness of a claim affects perceived credibility. Using an online experiment, participants will assess reports generated by others in controlled settings, with some exposed to incentives for being believed and others provided with gender information about the sender. We test whether people behave as Bayesians when assessing potential deception and whether incentives, gender, and belief priors systematically influence their evaluations. Our findings contribute to the understanding of truth perception, the strategic modification of statements based on audience incentives, and the mechanisms underlying the spread of misinformation.
External Link(s)

Registration Citation

Citation
Deryugina, Tatyana, Benjamin Ho and Olga Shurchkov . 2025. "Lies, Truths, and Perceptions." AEA RCT Registry. February 05. https://doi.org/10.1257/rct.15324-1.0
Sponsors & Partners
Experimental Details

Interventions

Intervention(s)
The study involves an online experiment conducted via Qualtrics and Prolific, where participants evaluate the truthfulness of reports made by others. The experiment consists of three main parts:

Truthful Self-Reporting Task (Part 1): Participants (senders) complete a task (either a random number task or a matrix task) and report their performance.
Lie Detection Task (Part 2): Other participants (receivers) assess the accuracy of reports from Part 1, sometimes with additional information such as gender cues. Receivers also express their confidence and wager potential earnings on their assessments.
Survey and Beliefs (Part 3): Participants provide demographic information and respond to incentivized questions about general misreporting patterns.
The intervention includes:

Two experimental tasks (Numbers Task and Matrix Task)
Four treatment conditions per task:
No Incentive to Persuade (NIT) & No Gender Information (NGT)
Incentive to Persuade (IT) & No Gender Information (NGT)
No Incentive to Persuade (NIT) & Gender Information (GT)
Incentive to Persuade (IT) & Gender Information (GT)
Intervention Start Date
2025-02-05
Intervention End Date
2026-02-25

Primary Outcomes

Primary Outcomes (end points)
Primary Endpoints:
Truth Perception Accuracy: The probability that receivers correctly assess whether a report is true or false.
Earnings Maximization: Whether receivers make truth assessments in a way that maximizes their expected earnings.
Effect of Report Extremeness on Credibility: Whether more extreme claims (e.g., reporting a maximum value) are perceived as less truthful.
Effect of Incentives on Lying: Whether senders report more truthfully when their payment depends on being believed.
Primary Outcomes (explanation)
see attachment

Secondary Outcomes

Secondary Outcomes (end points)
Confidence and Betting Behavior: How much participants stake on their evaluations and whether they adjust this behavior under different treatments.
Gender Effects on Believability: Whether gender cues influence the perceived credibility of reports.
Receiver’s Own Lying Behavior: Whether a participant’s own tendency to lie affects their likelihood of doubting others.
Cognitive Processing and Lie Detection: The relationship between contemplation time and lying behavior.
Social and Risk Preferences: How demographic and preference-based factors correlate with lying behavior and lie detection accuracy.
Altruism in Lie Detection: Whether receivers avoid marking reports as false in cases where doing so reduces the sender’s earnings.
Secondary Outcomes (explanation)
see attachment

Experimental Design

Experimental Design
Within-subjects variation: Each receiver evaluates multiple reports of varying extremeness.
Between-subjects variation: Participants are randomized into one of four treatment conditions (incentive vs. no incentive, gender information vs. no gender information).
Attention Checks: To ensure data quality, an explicit attention check will be implemented, and subjects failing it will be removed.

(see attachment)
Experimental Design Details
Not available
Randomization Method
qualtrics survey
Randomization Unit
individual
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
2000 online subjects
Sample size: planned number of observations
2000
Sample size (or number of clusters) by treatment arms
1000 in the effort task, 1000 in the random number task. also within each task, we randomized information provided about gender, and incentives so 250 for each arm of the 2x2 treatment design.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
The probability of evaluating a report as true in pilot studies was ~60%. With 2,000 participants evaluating three reports each: The 95% confidence interval for this probability is 58.8% to 61.2%. When analyzing the two tasks separately: Effective sample size per task is 1,000 participants, confidence interval 58.2% to 61.8%.
Supporting Documents and Materials

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

Request Information
IRB

Institutional Review Boards (IRBs)

IRB Name
UIUC
IRB Approval Date
2024-02-14
IRB Approval Number
RB23-0365
Analysis Plan