Support for affirmative action

Last registered on October 05, 2022
View Trial History

Pre-Trial

Trial Information

General Information

Title
Support for affirmative action
RCT ID
AEARCTR-0010130
Initial registration date
September 29, 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
October 05, 2022, 11:41 AM EDT

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

Locations

Country
United States of America
Region
Texas

Primary Investigator

Name
Leonardo Bursztyn
Affiliation
University of Chicago
Contact Primary Investigator

Other Primary Investigator(s)

PI Name
Alexander Cappelen
PI Affiliation
Norwegian School of Economics
Contact Investigator
PI Name
Bertil Tungodden
PI Affiliation
Norwegian School of Economics
Contact Investigator
PI Name
Alessandra Voena
PI Affiliation
Stanford University
Contact Investigator
PI Name
David Yanagizawa-Drott
PI Affiliation
University of Zurich
Contact Investigator

Additional Trial Information

Status
In development
Start date
2022-09-29
End date
2022-10-31
Keywords
Gender, Other
Additional Keywords
Social norms, Misperceptions
JEL code(s)
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
What people perceive as the norms or attitudes in society affects important behavior. The authors document misperceptions about gender norms with nationally representative data, covering 80% of the world population. In this study, we are additionally interested in studying beliefs about more direct peer groups, such as fellow citizens at the state level or coworkers. Do individuals also hold incorrect beliefs about their more direct peer groups? Is information about national support levels relevant for how individuals form beliefs about their peers?
External Link(s)

Registration Citation

Citation
Bursztyn, Leonardo et al. 2022. "Support for affirmative action." AEA RCT Registry. October 05. https://doi.org/10.1257/rct.10130-1.0
Sponsors & Partners
Experimental Details

Interventions

Intervention(s)
Half of the participants in our trial will be shown information about the beliefs of others regarding affirmative action. (See Experimental Design)
Intervention Start Date
2022-09-29
Intervention End Date
2022-10-07

Primary Outcomes

Primary Outcomes (end points)
Our primary outcome of interest is subjects' misperceptions of gender norms.
Primary Outcomes (explanation)
We define misperceptions as the difference between perceived – and actual gender norms. To elicit actual gender norms, we ask subjects whether they agree with prioritizing women when hiring for leadership positions. We measure perceived gender norms by asking about subjects to guess the percentage of the respective peer group that support the previous statement.

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
We will recruit subjects currently residing in Texas. Our study consists of three parts:
First, subjects will be asked whether or not they agree with a statement advocating affirmative action. They will also be asked about their perception of support for affirmative action among US-Americans.
Subsequently, half of the respondents simply confirm their answers from the previous block. Half of the respondents confirm their answers from the previous part and additionally receive information about the true support among US-Americans, as measured through nationally representative data the authors collected prior to this study.
Finally, subjects will be asked about their perception of support for affirmative action among people living in Texas and among their coworkers.
Experimental Design Details
Our partner company Prolific will recruit 500 participants currently residing in Texas, United States. Our sample will include 50% women and 50% men. The subjects complete an online survey created with the software Qualtrics. We will pool data from a pilot study with nearly identical wording and a sample size of N=200 for robustness regressions.

The statement we are interested in is the following: "The government and companies should give priority to women when hiring for leadership positions."

The survey begins by eliciting subjects’ private (first order) beliefs about this statement (agree/disagree). Subsequently, subjects are asked to guess the “percentage of [men/women] in the United States” that would agree with the statement (second order beliefs). This latter question is incentivized: “We asked the same question you were just asked to a random sample in the US that is nationally representative. In the second part of this survey, we would like to know what you believe these people answered. If your guess is close to the truth (within 2 percentage points), you will earn an additional $1 USD per question.”

Our experimental intervention occurs after the elicitation of second order beliefs. Half of the respondents will be informed about the true share of US Americans that agreed with the statement when interviewed for the nationally representative survey. Concretely, subjects in the treatment condition will be told: “You guessed that [subject’s answer] % of [MEN/WOMEN] agree with the statement that government and companies should give priority to women when hiring for leadership positions. According to the nationally representative survey, the actual share of [men/women] that agree with this statement is [38/45] %.”

After the intervention, we elicit subjects’ beliefs about support for the statement among individuals living in Texas and among their coworkers. Specifically, we ask participants the following questions: “Please guess: What percentage of [MEN in Texas/MEN you work with/WOMEN in Texas/WOMEN you work with] would say that they agree with the following statement? ‘The government and companies should give priority to women when hiring for leadership positions.’”
We randomize whether subjects will be asked about Texans or their coworkers first. Additionally, we randomize how we introduce the two questions asking about Texans. Half of the subjects will be introduced to the questions as follows: “In this part of the survey, we will ask you a few questions about your perceptions of what people living in Texas think.” The other half of subjects will be introduced as follows: “We asked the same question you were just asked to a random sample of men and women in Texas. In this part of the survey, we will ask you a few questions about your perceptions of what people living in Texas think. If your guess is close to the truth (within 2 percentage points), you will earn an additional $1 USD per question.” The data from which we calculate the truth is the same used to incentivize the very first elicitation of second order beliefs.

We hypothesize that:
- Misperceptions about support for affirmative action among Texans are statistically and economically significantly different from zero.
- Misperceptions about support for affirmative action among US citizens correlate positively with misperceptions about support among Texans.
- Subjects in the treatment condition take the new information into account when forming beliefs about Texans or their coworkers. That is, we expect the subjects' guesses about support for affirmative action in Texas/among co-workers to be significantly different between the two groups (information/no information).
Randomization Method
Randomization will be done using our survey software Qualtrics.
Randomization Unit
Randomization will be conducted at the individual level. All randomizations we use throughout the survey are done independently of each other.
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
500 individuals (i.e. no clusters)
Sample size: planned number of observations
500 individuals
Sample size (or number of clusters) by treatment arms
250 individuals in the control group, 250 individuals in the (information provision) treatment group
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Supporting Documents and Materials
IRB

Institutional Review Boards (IRBs)

IRB Name
Social and Behavioral Sciences Institutional Review Board at the University of Chicago
IRB Approval Date
2022-09-19
IRB Approval Number
IRB22-1356
Analysis Plan

Post-Trial

Post Trial Information

Study Withdrawal

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

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials