x

We are happy to announce that all trial registrations will now be issued DOIs (digital object identifiers). For more information, see here.
Endogenous and exogenous reputation in the one-shot, sequential prisoner dilemma.
Last registered on November 06, 2019

Pre-Trial

Trial Information
General Information
Title
Endogenous and exogenous reputation in the one-shot, sequential prisoner dilemma.
RCT ID
AEARCTR-0004937
Initial registration date
Not yet registered
Last updated
November 06, 2019 2:44 AM EST
Location(s)

This section is unavailable to the public. Use the button below to request access to this information.

Request Information
Primary Investigator
Affiliation
Université Libre de Bruxelles
Other Primary Investigator(s)
PI Affiliation
Université Libre de Bruxelles
PI Affiliation
Université Libre de Bruxelles
Additional Trial Information
Status
In development
Start date
2019-11-07
End date
2020-05-05
Secondary IDs
Abstract
The aim of the experiment is to test the effectiveness of reputation mechanisms to promote cooperation in one-shot, anonymous interactions.
In a sequential prisoner dilemma, we vary whether past behavior of the second mover is disclosed to the first mover. We also vary whether the disclosure is exogenous or endogenous (i.e. a choice by the second mover).
External Link(s)
Registration Citation
Citation
Kirchsteiger, Georg, Tom Lenaerts and Rémi Suchon. 2019. "Endogenous and exogenous reputation in the one-shot, sequential prisoner dilemma.." AEA RCT Registry. November 06. https://doi.org/10.1257/rct.4937-1.1.
Experimental Details
Interventions
Intervention(s)
We experimentally study the impact of different types of reputation on cooperation in a one shot, (modified) sequential prisoner dilemma.
We focus on the reputation of the second mover. Between conditions, we vary whether the reputation is noisy and whether its disclosure is endogenous, i.e. chosen by the second mover.



Intervention Start Date
2019-11-07
Intervention End Date
2020-05-05
Primary Outcomes
Primary Outcomes (end points)
The main variables of interest are:

The level of cooperation, by both first and second movers.
The decision of first movers to enter into the prisoner dilemma.
The decision of second movers to disclose information about their past choices.
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
We experimentally study the impact of reputation on cooperation in a one shot, modified sequential prisoner dilemma. We modify the prisoner dilemma by allowing the first mover to opt-out of the interaction, and secure instead a sure amount.

In each session, participants play the sequential prisoner dilemma for 30 rounds. Roles ("first mover" and "second mover") are fixed. We rematch participants randomly between each rounds.

In total, we plan to run 7 experimental conditions:

We implement a condition in which no reputation mechanism is implemented. We use it to establish a baseline level of cooperation.

In addition, we implement a 2 by 3 between-subjects factorial design:

1- In the first dimension, we vary whether the disclosure of information about the past choices of the second mover is disclosed exogenously or endogenously to the first mover. In the "exogenous" conditions, the disclosure of information about the past choices of the second mover is totally exogenous. In the "endogenous" conditions, the second mover decides whether he wants to disclose information about his past choices.

2- In the second dimension, we vary the noisiness of reputation. In the exogenous reputation conditions, noise is introduced by means of a probability that the information about the past choices of the second mover is disclosed. In the endogenous reputation conditions, noise is introduced by means of a probability that the information about the past choices of the second mover is disclosed conditional on the choice of the second mover to disclose this information. There are three levels of noise: no noise (the probability on disclosure is one), low noise (the probability of disclosure is 90%) and high noise (the probability of disclosure is 60%).

Experimental Design Details
Not available
Randomization Method
Computer randomization.
Randomization Unit
Conditions are implemented at the session level.
Was the treatment clustered?
Yes
Experiment Characteristics
Sample size: planned number of clusters
420 individuals in total, 210 in each role.
Sample size: planned number of observations
12600 observations in total (30 rounds * 420 individuals), 6300 in each role (30 rounds * 210 individuals).
Sample size (or number of clusters) by treatment arms
21 sessions (three in each condition). In each condition, 30 individuals in each role.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Given a target power of 80% and a significance level of 5%, we are able to detect a treatment effect corresponding to 15 percentage point. (i.e. cooperation increasing or decreasing by 15 percentage points between treatments). Details are given in the supplementary documents.
Supporting Documents and Materials

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

Request Information
IRB
INSTITUTIONAL REVIEW BOARDS (IRBs)
IRB Name
IRB Approval Date
IRB Approval Number