Back to History Current Version

Faces or Facts? Experimental Evidence on Time Pressure, Information Costs, and Gender Bias in Electoral Choice

Last registered on January 06, 2026

Pre-Trial

Trial Information

General Information

Title
Faces or Facts? Experimental Evidence on Time Pressure, Information Costs, and Gender Bias in Electoral Choice
RCT ID
AEARCTR-0017574
Initial registration date
December 30, 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
January 06, 2026, 7:13 AM EST

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

Last updated
January 06, 2026, 7:41 AM EST

Last updated is the most recent time when changes to the trial's registration were published.

Locations

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

Request Information

Primary Investigator

Affiliation
Universidad de O'Higgins

Other Primary Investigator(s)

PI Affiliation

Additional Trial Information

Status
In development
Start date
2026-01-05
End date
2026-06-30
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
This project seeks to analyze how physical attributes, such as perceived dominance, and the policy proposals of candidates interact in electoral decisions, considering gender and information conditions (time pressure and access costs) as factors that structure biases and stereotypes in political behavior.

The research stems from evidence that when voters face cognitive or informational limitations, they tend to use heuristic shortcuts, such as physical appearance, to simplify their decision. Previous studies demonstrate that traits such as attractiveness, facial dominance, and perceived emotions influence the evaluation of candidates, and that these effects can differ between men and women, reproducing leadership and gender stereotypes.

The study proposes four main hypotheses: (1) under time pressure, the influence of physical attributes increases; (2) when accessing information is costly, voters prioritize visual cues; (3) female candidates are evaluated more by their appearance than male candidates, especially under conditions of low information; and (4) there is a tendency to value proposals from candidates of the same gender more highly.

The research uses a quantitative experimental design divided into two complementary phases. In the first phase (pretesting), the measurements of physical attributes obtained using the Face++ application will be validated through human evaluations of a set of 108 real photographs of mayoral candidates in Colombia. Each participant will evaluate 16 photographs on a Likert scale of 1 to 7, indicating the degree to which they consider the person portrayed to be dominant. These evaluations will allow for comparing the consistency of the algorithmic indicators with human perception, ensuring the validity of the visual stimuli that will be used in the experimental phase. Approximately 102 participants will be required to obtain at least 1,620 valid evaluations.

The second phase corresponds to the implementation of a conjoint experiment, a design widely used in the social sciences to identify the relative weight of multiple attributes in individual decisions. In this study, 444 participants (men and women aged 18 to 65, neither residents nor citizens of Colombia) will be presented with a series of candidate profile pairs containing randomized information about their gender, age, ideology, physical attractiveness, and policy proposals. The profiles will include a photograph (obtained from the ballot papers of the last four local elections in Colombia) and a brief summary of ideas compiled from actual campaign manifestos and condensed using ChatGPT.
The sample will be divided into three groups: a control group, a group subjected to time pressure, and a group with information access costs. Through this experimental manipulation, we will estimate how information conditions modify the weight of physical and gender attributes in electoral decisions. All participants will sign informed consent forms, the anonymity of responses will be guaranteed, and the data will be used exclusively for academic purposes.

The study aims to understand whether voters rely more on appearance when they feel rushed or when obtaining information is costly, and whether these effects differ for male and female candidates and voters. By doing so, the research seeks to shed light on how information constraints and gender-related stereotypes shape political decision-making. The findings will contribute to a better understanding of how real-world voting conditions influence electoral choices and democratic representation.
External Link(s)

Registration Citation

Citation
Canales, Andrea and Alejandra Marín. 2026. "Faces or Facts? Experimental Evidence on Time Pressure, Information Costs, and Gender Bias in Electoral Choice." AEA RCT Registry. January 06. https://doi.org/10.1257/rct.17574-1.1
Sponsors & Partners

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

Request Information
Experimental Details

Interventions

Intervention(s)
The study implements a randomized online survey experiment. Participants are asked to make a series of voting choices between pairs of hypothetical mayoral candidates. Each candidate profile includes a photograph and a brief summary of policy positions. Candidate attributes, such as gender, age category, political ideology, and facial characteristics related to perceived dominance, are randomly varied across profiles.
Participants are randomly assigned to one of three experimental conditions:
- Control: Participants view candidate profiles with photographs and brief policy descriptions and make voting choices without additional constraints.
- Time Pressure: Participants view the same candidate profiles but must make each voting decision within a fixed and limited amount of time, indicated by a visible countdown timer.
- Information Cost: Participants may access more detailed policy information for each candidate by clicking an additional option. Accessing this information requires completing an extra step, making information more costly to obtain.
All participants complete the same number of choice tasks. After completing the voting tasks, participants answer short follow-up questions regarding perceived time pressure, perceived information costs, and their decision-making experience. No real political actors are involved, and the intervention does not include persuasion or behavioral manipulation beyond the randomized presentation of decision environments
Intervention Start Date
2026-03-01
Intervention End Date
2026-04-11

Primary Outcomes

Primary Outcomes (end points)
1. Voting Choice:
The main outcome is the participant’s voting choice in each paired candidate comparison. This outcome is recorded as a binary indicator identifying which candidate profile is selected in each choice task.
Attribute-Level Choice Effects
Based on the observed voting choices, the study will estimate the effect of candidate attributes—such as gender, age category, political ideology, and facial dominance—on the probability that a candidate is selected.
Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
The study uses a randomized online survey experiment with a between-subjects design at the participant level and a within-subjects design at the choice-task level. Participants are randomly assigned to one of three experimental conditions: a control condition, a time pressure condition, or an information cost condition.
All participants complete a series of paired candidate choice tasks. In each task, participants choose between two hypothetical mayoral candidates. Each candidate profile includes a photograph and a brief summary of policy positions. Candidate attributes, such as gender, age category, political ideology, and facial characteristics related to perceived dominance are randomly assigned across profiles and across choice tasks.
The experimental conditions differ only in the decision environment. In the control condition, participants make choices without additional constraints. In the time pressure condition, participants must make each choice within a fixed and limited amount of time, indicated by a visible timer. In the information cost condition, participants may access more detailed policy information by completing an additional step, making information more costly to obtain.
The order of choice tasks and the pairing of candidate profiles are randomized. All participants complete the same number of choice tasks. Data collection is conducted online using an external survey provider, and participants are not exposed to real political campaigns or actors.
Experimental Design Details
Not available
Randomization Method
Qualtrics
Randomization Unit
Randomization is conducted at multiple levels.
At the participant level, individuals are randomly assigned to one of three experimental conditions: control, time pressure, or information cost.
At the task level, candidate profiles and their attributes (including gender, age category, political ideology, and facial dominance) are randomly assigned across choice tasks within each participant. The order of tasks and the pairing of candidate profiles are also randomized at the individual level.
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
444 individual participants
Sample size: planned number of observations
The planned sample includes 444 individual participants. Each participant completes 8 paired candidate choice tasks, resulting in a total of 3,552 choice-task observations (evaluations).
Sample size (or number of clusters) by treatment arms
Control: 148 participants, yielding 1,184 choice-task observations
Time Pressure: 148 participants, yielding 1,184 choice-task observations
Information Cost: 148 participants, yielding 1,184 choice-task observations
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Approximately 4–5 percentage points in the probability of choosing a profile (binary choice), assuming SD = 0.50 and clustering at the participant level.
IRB

Institutional Review Boards (IRBs)

IRB Name
Universidad de O'Higgins
IRB Approval Date
2025-12-03
IRB Approval Number
086-2025