Adapting Economic Games to Personalize Privacy Nudges

Last registered on April 23, 2024
View Trial History

Pre-Trial

Trial Information

General Information

Title
Adapting Economic Games to Personalize Privacy Nudges
RCT ID
AEARCTR-0012875
Initial registration date
February 08, 2024

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
February 14, 2024, 12:06 PM EST

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

Last updated
April 23, 2024, 6:20 PM EDT

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

Locations

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

Request Information

Primary Investigator

Name
Qingyi Wang
Affiliation
University of Michigan
Contact Primary Investigator

Other Primary Investigator(s)

PI Name
Yan Chen
PI Affiliation
University of Michigan
Contact Investigator
PI Name
Eytan Adar
PI Affiliation
University of Michigan
Contact Investigator

Additional Trial Information

Status
In development
Start date
2023-09-01
End date
2024-06-30
Keywords
Behavior
Additional Keywords
JEL code(s)
D83, D91
Secondary IDs
NSF Grant 2209507
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
Modern social communication systems–ranging from email to social media systems–present a dizzying number of decisions for users. Privacy configurations, when not hidden by social media companies, are opaque. Individuals sometimes are also not aware of the externality that their sharing decisions can have to others. Thus, it is often hard for individuals to react or behave in ways that model the personal behaviors or are communally advantageous. Personalized recommendations, interfaces, interventions or nudges can help but implementing these requires an understanding end-user preferences. Our research seeks to tackle this challenge by modeling individual preferences through the use of economic games, both in a neutral context and in specific scenarios. Simultaneously, we will collect user preferences in actual social and communication systems. We propose to connect the game-modeled behaviors with real-world preferences. Doing so will allow us to create interventions that can help align interface settings with real preferences or can nudge towards better decisions.
External Link(s)

Registration Citation

Citation
Adar, Eytan, Yan Chen and Qingyi Wang. 2024. "Adapting Economic Games to Personalize Privacy Nudges." AEA RCT Registry. April 23. https://doi.org/10.1257/rct.12875-1.3
Sponsors & Partners

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

Request Information
Experimental Details

Interventions

Intervention(s)
We exogenously vary the correlation coefficient between one's own information and their match's information, as well as the payment provided for sharing the information, to observe how participants change their information-sharing decisions in both neutral contexts and real life scenarios.
Intervention Start Date
2024-01-24
Intervention End Date
2024-04-30

Primary Outcomes

Primary Outcomes (end points)
In the information sharing game, the primary outcome is the likelihood that participants select the best responses given by the realized payoff matrices.
In the friend decision making module. the primary outcome is the willingness to accept to announce answers elicited by the BDM method.
Primary Outcomes (explanation)
In the information sharing game, participants make binary sharing decisions, share or not share. We also elicit their beliefs about what their match will choose in each round, share or not share. We calculate the best responses in each round given their beliefs, using the realized payoff matrices. The primary outcome is a binary variable indicate whether participant's decision is consistent with the best response we calculate.

Secondary Outcomes

Secondary Outcomes (end points)
In the information sharing game, the secondary outcomes are (1) the likelihood that participants select the mutual best responses given by the realized payoff matrices; (2) payoffs; (3) efficiency
Secondary Outcomes (explanation)
The likelihood that participants select the mutual best responses given by the realized payoff matrices is an indicator of whether participants' and their match both choose best responses given their beliefs
Participants' payoffs in each round consists of two parts. One part is the sum of squared differences between each pair of their secret numbers and the robot's guesses. The other part is the payment offered by the robot for sharing their secret numbers.
Efficiency is calculated by (payoff - minimum payoff)/(maximum payoff - minimum payoff) in each round.

Experimental Design

Experimental Design
The experiment has three modules:
The first module is an information sharing game. We test a simplified version of the Acemoglu et al. (2022) model in neutral contexts.
The second module introduces friend decision making. We assign specific real-life scenarios based on a similar game structure as the first module.
Last, we elicit participants' risk/ambiguity/social preferences using standard economic games, and investigate whether participants' information-sharing decisions in the first two modules can be predicted by the game outcomes.
Experimental Design Details
Not available
Randomization Method
First, the two versions of the information sharing game (with or without the average payoff matrices) are randomly assigned to each experiment session.
Second, at the beginning of each experiment session, participants randomly draw a card to decide their participant label in the session.
Third, in the information sharing game, we do complete random rematching in each round using oTree, to make sure that participants are never matched with the same participant again.
Randomization Unit
The randomization of game versions is at the session level.
The randomization of participant labels and the random rematching in the game are at the individual level.
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
We plan to conduct 20 sessions of the experiment.
Sample size: planned number of observations
We plan to recruit 240 participants in total, where there are 12 in each session.
Sample size (or number of clusters) by treatment arms
Out of the 20 sessions, 10 sessions will be assigned with the version where participants see the average payoff matrices first, and 10 sessions will be assigned with the version where participants make their own decisions first.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Based on our pilot data, to observe a 0.2 difference in the likelihood of choosing the best response (where the likelihood is 0.8 in the rounds with the average payoff matrices, and the likelihood is 0.6 in the rounds without the average payoff matrices), with a standard deviation of 0.45, and the correlation coefficient between the two types of decisions to be 0.25, the minimum detectable effect size under a within-subject design is 61.
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
The Health Sciences and Behavioral Sciences Institutional Review Board (IRB-HSBS)
IRB Approval Date
2023-05-31
IRB Approval Number
HUM00236045
Analysis Plan

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

Request Information