Legitimacy And Social Desirability: MTurk Online Survey

How to cite

Fiorin, Stefano; Egorov, Georgy; Bursztyn, Leonardo. 2017. "Legitimacy And Social Desirability: MTurk Online Survey." AEA RCT Registry. AEARCTR-0002028. https://doi.org/10.1257/rct.2028-5.0

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

Stefano Fiorin* , Georgy Egorov , Leonardo Bursztyn

* Primary investigator

Registry ID
AEARCTR-0002028
DOI
https://doi.org/10.1257/rct.2028-5.0
First registered
Last updated
Status
Completed
Trial dates
2017-02-21 – 2017-03-21
Countries
United States of America (California, Hawaii, Maryland, Massachusetts, New York, and Vermont. Adding Connecticut, Delaware, Illinois, New Jersey, Rhode Island, Washington because there are not enough respondents from initial states.), United States of America ()
Registry record
http://www.socialscienceregistry.org/trials/2028

Abstract

We will recruit survey participants on MTurk to conduct an online survey. With this survey, we plan to study whether the legitimacy (in the sense of conformity to the law) of a certain political view also affects its social desirability.
First, all subjects will answer a few demographic questions. Second, they will be asked about their opinion on prohibiting the building of minarets. Third, they will be asked to participate in a dictator game where they will have to split $3 between them and another subject. They will be randomized into four groups. Subjects in the control group will only be told that the other subject is a 24-year-old male from Switzerland. Subjects in the "anti-minarets" treatment group will be also told that the other subject is against the building of minarets. Subjects in the "anti-minarets public support" will additionally be told that the 57.5% of Swiss voters is against minarets. Finally, subjects in the "anti-minarets legitimacy" will be additionally told that following the result of a referendum in 2009, building minarets is illegal in Switzerland.
Our main hypotheses are:
1) The donation rates will be reduced in the "anti-minarets" when compared to the donation rates in the control group.
2) The donation rates will be increased in the "anti-minarets legitimacy" when compared to the donation rates in the "anti-minarets public support" group.

In a follow-up experiment, we recruit participants through an online panel survey firm who previously self-identified to the firm as Democrats. We study how the 2016 US presidential election changed people’s interpretation of observed xenophobic expression. First, all participants answer a few demographic questions. Second, the participants (player 1) are asked to play a dictator game with a random participant to a previous study (player 2) that they are matched with. Players 1 are informed that in the other study player 2 authorized a donation to an anti-immigration organization. We cross-randomize whether we tell player 1 that: i) Trump or Clinton won the election in player 2’s area (respectively "Trump won" and "Clinton won" treatment); and ii) player 2’s donation was in private or publicly observable on a website shared with other participants from his/her area (respectively "private" treatment and 'public" treatment). Our hypothesis is that player 1 donation rates to player 2 will be higher in the "Trump won public" treatment than in the other three treatment conditions.

Experimental design

We will recruit survey participants on MTurk to conduct an online survey. With this survey, we plan to study whether the legitimacy (in the sense of conformity to the law) of a certain political view also affects its social desirability.
First, all subjects will answer a few demographic questions. Second, they will be asked what is their opinion on prohibiting the building of minarets. We will focus our analysis on participants who are against prohibiting the building of minarets. Third, they will be asked to participate in a dictator game where they will have to split $3 between them and another subject. They will be randomized into four groups. Subjects in the control group will only be told that the other subject is a 24-year-old male from Switzerland. Subjects in the "anti-minarets" treatment group will be also told that the other subject is against minarets. Subjects in the "anti-minarets public support" will additionally be told that 57.5% of Swiss voters is against minarets. Finally, subjects in the "anti-minarets legitimacy" will be also told that following the result of a referendum, building minarets is illegal in Switzerland.
Participants in the control group will be asked whether they think minarets are illegal in Switzerland, and what percentage of Swiss people they think are against minarets.

In a follow-up experiment, we recruit participants through an online panel survey firm who previously self-identified to the firm as Democrats. First, all participants answer a few demographic questions. Second, we randomize the participants in one of two groups. In the "Trump won" group we tell participants that a respondent from another survey comes from an area where Trump won the election. In the "Clinton won" group we tell participants that a respondent from another survey comes from an area where Clinton won the election. Then all subjects are presented with two anti-immigrant quotes and are asked to predict the share of voters in the other participant’s area that
they believed would agree with the quotes. Third, the participants (player 1) are asked to play a dictator game with a random participant to a previous study (player 2) that they are matched with. Players 1 are informed that in the other study player 2 authorized a donation to an anti-immigration organization. We cross-randomize with the Trump/Clinton won treatments whether we tell player 1 that player 2’s donation was in private or publicly observable on a website shared with other participants from his/her area (respectively "private" treatment and 'public" treatment).

Design details

Randomization method
The randomization will be done on Qualtrics, a website for conducting online surveys.
Randomization unit
The unit of randomization will be the individual subject.
Clusters
N/A
Observations
400 In the second experiment, 1,800
Arms
100 control group, 100 anti-minarets, 100 anti-minarets public support, 100 anti-minarets referendum In the second experiment: 450 Trump won public, 450 Trump won private, 450 Clinton won public, 450 Clinton won private
Primary outcomes
Dictator game decisions.

Materials

Cited by

No citation data available for this registration yet.

Keywords