Social and Self-Image: an Experimental Study

How to cite

Galbiati, Roberto; Henry, Emeric; Jaquemet, Nicolas. 2023. "Social and Self-Image: an Experimental Study." AEA RCT Registry. AEARCTR-0011186. https://doi.org/10.1257/rct.11186-1.0

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

Roberto Galbiati* , Emeric Henry , Nicolas Jaquemet

* Primary investigator

Registry ID
AEARCTR-0011186
DOI
https://doi.org/10.1257/rct.11186-1.0
First registered
Last updated
Status
In development
Trial dates
2023-04-11 – 2023-09-30
Countries
France ()
Registry record
http://www.socialscienceregistry.org/trials/11186

Abstract

While there is a growing literature in both economics and psychology on the consequences of image concerns on ethical and cooperative behaviors, self-image and social-image have not been studied jointly. The first aim of this study is to investigate the linkages between self and social image. We introduce theoretically, and aim to test empirically, a new channel that helps to explain how self-image can be strategically manipulated.

It is well-known that individuals strategically preserve their self-image by manipulating their beliefs, for instance by avoiding information (Dana, Weber and Kuang, 2007, Econ. Th.). This in turn can affect the way they judge others, through what we call the “coherence principle”. This coherence can be purely behavioral (what we refer to as behavioral coherence) and/or can be informational; i.e., if an individual ignores information to manipulate her self-image, the resulting moral judgement might be shifted by the resulting lack of information (what we refer to as informational coherence).

This experiment is designed to test the coherence principle. In order to do so we will put subjects in context where they have to judge others’ actions in different situations. In one treatment, subjects will have to judge the morality of others’ actions in situations where they do not face similar choices, while in another treatment subjects will have to judge others’ in a situation where they face similar choices. We will also introduce the informational channel in a separate set of experiments, in which observers will receive the judgement-relevant information without facing the choice.
This setting allows us to answer the following questions:
1. Does the judgement of other’s (pro-social or selfish) actions change when observers face similar decisions?
2. Do subjects anticipate that other’s judgement of their action depends on the possibility that observers have to take the same actions?
3. Does the option of accessing information on the action reinforce the coherence principle?

The experiment will focus on the decision to give a sum of money to an NGO.

Intervention

The experimental subjects will be asked to take donation decisions in favor of an NGO under different conditions about observability of their choices by other participants and different information conditions.

Donation decision
Before making the decision, the observed will be allocated to one of three experimental conditions:
T1 in which their actions are not observed;
T2 in which the observed are told their actions will be observed by individuals (the observers), who will evaluate their moral conduct and whose rating will subsequently be transmitted to them
T3 same setting as in T2 except that the observed are told in addition that the observers will play exactly the same game as them before evaluating their moral conduct

Observers
The observers are accordingly allocated to either of the two treatments (there are no observers in T1):
- in T2 they simply have to rate the behavior of the observed on a morality scale (Passive observers).
- in T3 they play the same game as the observed before rating the behavior of the observed (Active observers)
In both treatments, the observers will have to rate donation choices on a Likert scale and can also leave written messages.

Information
We will have two set of experiments with respect to the information about the NGO.
- The ‘non-information’ treatments: observers and observed take their decisions without knowing the NGO’s identity

- The ‘information’ treatments (denoted with letter b): The identity of the NGO is not known at the beginning of the game. However, subjects have the option to discover it. In practice they have a screen where the click yes or no to the question “I want to learn the identity of the NGO to be printed at the bottom of the screen”:
o For the observed In T1b-T3b the participants have the option to view the NGO’s identity
o For the observers, in T2b and T3b, the observers have the option to view the NGO’s identity before they evaluate. In T3b they also can learn the NGO’s identity before they donate
o In every treatment, both observers and observed know that the other party (observed and observer respectively) have the option to discover the NGO identity

Post decision/observation phase
After the decision/evaluation phase we will ask some questions on:
1. Individual self-evaluation of their actions (donate-not donate)
2. Their evaluation of the public utility of NGOs
3. Psychometric questionnaires on self-esteem and concern for appropriateness

Experimental design

Recruitment and Pool of participants
The experiment will be run online. We will use the services of Foule Factory, a firm that gives us access to a pool of participants, the experiment will be programmed on Qualtrics. Subjects first respond to a survey on socio-demographic characteristics. After the survey, subjects will be divided into two groups, the “observers” and the “observed”. The observed receive an endowment and are asked to decide whether they want to keep the money for themselves, or give it up and trigger a donation to an NGO (whose identity will be revealed at the end of the experiment). This choice will be implemented with a 1/20 probability.
Donation decision
Before making the decision, the observed will be allocated to one of three experimental conditions:
T1 in which their actions are not observed;
T2 in which the observed are told their actions will be observed by individuals (the observers), who will evaluate their moral conduct and whose rating will subsequently be transmitted to them
T3 same setting as in T2 except that the observed are told in addition that the observers will play exactly the same game as them before evaluating their moral conduct

Observers
The observers are accordingly allocated to either of the two treatments (there are no observers in T1):
- in T2 they simply have to rate the behavior of the observed on a morality scale (Passive observers).
- in T3 they play the same game as the observed before rating the behavior of the observed (Active observers)
In both treatments, the observers will have to rate donation choices on a Likert scale and can also leave written messages.

Information
We will have two set of experiments with respect to the information about the NGO.
- The ‘non-information’ treatments: observers and observed take their decisions without knowing the NGO’s identity

- The ‘information’ treatments (denoted with letter b): The identity of the NGO is not known at the beginning of the game. However, subjects have the option to discover it. In practice they have a screen where the click yes or no to the question “I want to learn the identity of the NGO to be printed at the bottom of the screen”:
o For the observed In T1b-T3b the participants have the option to view the NGO’s identity
o For the observers, in T2b and T3b, the observers have the option to view the NGO’s identity before they evaluate. In T3b they also can learn the NGO’s identity before they donate
o In every treatment, both observers and observed know that the other party (observed and observer respectively) have the option to discover the NGO identity
Post decision/observation phase
After the decision/evaluation phase we will ask some questions on:
1. Individual self-evaluation of their actions (donate-not donate)
2. Their evaluation of the public utility of NGOs
3. Psychometric questionnaires on self-esteem and concern for appropriateness

Payment
We communicate to subjects the evaluation they have received by the observers, we reveal the identity of the NGO to whom the donation will go and we proceed to the payment online (via the platform). The NGO will send a certificate of donation to each participant that have decided to donate.

Design details

Randomization method
random allocation to treatments (rotation done by the computer)
Randomization unit
Individual
Clusters
2400 individuals
Observations
2400 individuals
Arms
300 per treatment arm
Primary outcomes
1. The first part will study evaluations of subjects actions by other subjectis Question 1: on average how do evaluations of action give vs keep compare? We use all observer treatments to perform analysis. We expect the action keep to be more harshly judged Question 2: Do observers rate more harshly the action 'keep 'when they are inactive? Comparison between T2 and T3 for observers. The coherence principle supposes that those who are active, and found excuses to justify their actions, are then more likely to apply these excuses when judging others Question 3: Does the gap between evaluations of active vs inactive grows in the information treatments? Here we use treatments T1,T2,T1b and T2b and we compare the gap in evaluations between T1 and T2 to the gap between T1b and T2b. The informational channel (the fact active players are less likely to look up the name of the NGO reinforces the coherence principle
Secondary outcomes
Secondary outcomes are the excuses used by those who don't donate and the evaluations of the actions of the others

Cited by

No citation data available for this registration yet.

Keywords