Back to History Current Version

Apologies in Public Goods Game

Last registered on January 15, 2019

Pre-Trial

Trial Information

General Information

Title
Apologies in Public Goods Game
RCT ID
AEARCTR-0003567
Initial registration date
November 15, 2018

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
November 17, 2018, 10:05 PM EST

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

Last updated
January 15, 2019, 5:05 AM EST

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

Locations

Primary Investigator

Affiliation
University of Warwick

Other Primary Investigator(s)

PI Affiliation
University of Warwick

Additional Trial Information

Status
On going
Start date
2018-11-01
End date
2020-11-01
Secondary IDs
Abstract
The project aims to examine the effect of apologies on coordination and cooperation, the underlying mechanism and its efficiency, in a public
goods experimental game.
External Link(s)

Registration Citation

Citation
Yeo, Jonathan and SHI ZHUO. 2019. "Apologies in Public Goods Game." AEA RCT Registry. January 15. https://doi.org/10.1257/rct.3567-2.0
Former Citation
Yeo, Jonathan and SHI ZHUO. 2019. "Apologies in Public Goods Game." AEA RCT Registry. January 15. https://www.socialscienceregistry.org/trials/3567/history/40209
Experimental Details

Interventions

Intervention(s)
Intervention Start Date
2018-11-01
Intervention End Date
2020-11-01

Primary Outcomes

Primary Outcomes (end points)
Average group contribution in treatment groups compared to control groups;
Dynamics of individual apologies and contribution in treatment and control groups;
Dynamics of individual estimation of other group members' contribution and evaluation of other group members' altruism in treatment and control groups
Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)
Individual characteristics and correlation with primary outcomes;
Individual contribution preference in one-shot public goods games and correlation with primary outcomes
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
An experimental lab setting will be used for the research. At the beginning of the session, each participant will be randomly assigned to 1 of the 3 treatments: control, public apologies, private apologies. Participants will be randomly assigned to groups of 4 and play a standard repeated public goods game for 16 periods. In the treatment groups, participants will be able to send apology messages to group members after contributions revealed. We also plan to conduct three extra treatments with regrouping in each period to further explore the mechanism.
Experimental Design Details
An experimental lab setting will be used for the research. Subjects will be recruited via the Sign-up System for paid research at the University of Warwick. The computerized experiment is programmed using o-tree. Participants attend the anonymous experiment using the computers in the Economics Lab at University of Warwick. There will be 20 participants in each session. //

At the beginning of the session, each participant will be randomly assigned to 1 of the 3 treatments: control, public apologies, private apologies. Regardless of the treatment, a participant will go through Section 1, Section 2, and a post-questionnaire. //

In section 1, participants will play a repeated public goods game for 16 periods with the same group members. In the control treatment, there are 2 stages in each period: in the decision stage, each participant needs to decide how many coins to contribute to the group project, estimate how many coins each other group member will contribute to the group project, and on a 7-point scale to what extent each other group member care about others; in the results stage, contribution of each group member is revealed to each participant; at the end of each period, each participant learn about their earnings from the public goods game and estimation in that period. //

In the two apologies treatments, the decision stage is the same as the control treatment; however, in the results stage, below the contribution results, participants are given the opportunity to send apology messages: in public apologies treatment, participants are asked whether they want to send a message saying "I am sorry" to the group, while in private apologies treatment, participants are asked whether they want to send a message saying "I am sorry" to each other group member respectively; there is a third stage in the two apologies treatments, where apology messages are shown: in public apologies, participants can see whether each of the group members chose to send a public apology, while in private apologies, participants can only see whether each of the other group members chose to send a private apology to him/her. Same as the control treatment, at the end of each period, each participant learn about their earnings from the public goods game and estimation in that period. //

After the instructions, there is 1 period of Practice Round for the participants to better understand the rules. //

In Section 2, participants in the same treatment will be randomly assigned to a new group of 4. The design of Section 2 follows the P-experiment in Fischbacher & Gächter (2010) while the parameters of the public goods game is the same as in Section 1. Each participant need to choose an unconditional contribution, fill out the conditional contribution table, then they will see the contribution of each group member in Section 2. Afterwards, they will see a summary of their earnings in each period in both Section 1 and Section 2. 4 periods in Section 1 will be randomly selected for final payment along with earnings in Section 2.//

Then participants will be asked to fill out a post-questionnaire via a Qualtrics link. The questionnaire includes questions on basic demographic information, open-ended question on their decisions and experiences in the experiment, scale questions to measure their risk-preference, social preferences, gulit and shame pronness, etc. Participants will be paid a separate fixed amount completion fee for completing the survey. //

After the survey, participants will see their total earnings (show-up fee+earnings from 4 random periods in section 1+earnings from section 2+questionnaire completion fee), be paid in cash and leave. //

Details of the experiment can be found in the attached instructions given to participants in the experiment and the post-questionnaire is also attached. //

We also plan to conduct three extra treatments in order to explore the mechanism. The only difference between the three extra treatments and the three main treatments will be in the extra treatments, in each period, participants in the same treatment will be randomly assigned to groups of 4 again, so they are no longer interacting with the same group members over the 16 periods. //

As a next step, we also plan to conduct experiments adding punishment to all the current treatments to investigate the interaction between apologies and punishment in public goods game.
Randomization Method
All randomizations are done via the python-based otree computer program.
Randomization Unit
Individual
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
312 to 432 participants
Sample size: planned number of observations
312 to 432 participants
Sample size (or number of clusters) by treatment arms
52 to 72 control (13 to 18 groups), 52 to 72 public apologies (13 to 18 groups), 52 to 72 private apologies (13 to 18 groups),
52 to 72 control with regrouping (13 to 18 groups), 52 to 72 public apologies with regrouping (13 to 18 groups), 52-72 private apologies with regrouping(13 to 18 groups)
For regrouping treatments, we might do 24 to 40 for both private and public apologies instead, or just do one of the two apologies, depending on the available funding.

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
The closest reference we can find for our main outcome - effect of apologies on contribution in public goods game - is the following paper. Reference: Haruvy, E., Li, S. X., McCabe, K., & Twieg, P. (2017). Communication and visibility in public goods provision. Games and Economic Behavior, 105, 276-296.; Effect size=(M2 - M1) /√((SD1^2 + SD2^2) ⁄ 2); MPCR=0.3: (76.8-6.7)/√((31.3^2 + 12.7^2) ⁄ 2)=2.08 SD; MPCR=0.75: (96.5-13.9)/√((6^2 + 18.6^2) ⁄ 2)=4.23 SD; MPCR in our public goods game=0.5, so we take the average of their effect sizes = 3.15 SD; Since our game is different from the game used in the paper and more abstract, our expected effect size might be much lower than that. We use effect size of 1 SD for power calculation. With an effect size of 1 SD, a sample size of >35 is sufficient to obtain power of 0.8 and alpha=0.05
Supporting Documents and Materials

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

Request Information
IRB

Institutional Review Boards (IRBs)

IRB Name
University of Warwick Economics Departmental Ethical Approval officer
IRB Approval Date
2018-10-25
IRB Approval Number
N/A
Analysis Plan

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

Request Information

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