Behavioral Experiments in Data Privacy

How to cite

Yagnaraman, Dhwani; Singh, Jaspreet; Vancel, James. 2020. "Behavioral Experiments in Data Privacy." AEA RCT Registry. AEARCTR-0006047. https://doi.org/10.1257/rct.6047-2.1

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

Dhwani Yagnaraman* , Jaspreet Singh , James Vancel

* Primary investigator

Registry ID
AEARCTR-0006047
DOI
https://doi.org/10.1257/rct.6047-2.1
First registered
Last updated
Status
Completed
Trial dates
2020-03-31 – 2020-07-15
Countries
India (), Kenya ()
JEL codes
C91
Registry record
http://www.socialscienceregistry.org/trials/6047

Abstract

With the penetration of technology into new markets, there are a number of questions around data privacy that have been insufficiently studied. Moreover, many of the core issues in data privacy are behavioral in nature, such as the intention-action gap and hyperbolic discounting manifested by consumers. We will study two key aspects of data privacy behaviors: 1) How to make consumers more privacy-conscious through different dissemination techniques of privacy policy agreements and streamlining the consent process to make it more effective and 2) Privacy as a business advantage for industry through exploration of privacy disclosure, control over data, and ways of sharing data. This study incorporates an experimental approach. We test people’s revealed preferences for sharing data through lab games programmed on computers and understand their stated preferences through vignettes and survey questions. Our results will be used to demonstrate methods that make users engage more with consent and privacy policies and allow businesses to benefit from more aware consumers and improved data sharing practices.

Intervention

The interventions of this study consisted of visual cues or primes, which accompanied a standard privacy policy. These primes varied in the number of units and amount of visual and written content they contained. All the interventions appeared at the same point during the experiment for all respondents. The interventions can be grouped into five main categories: star ratings, usage case, costly signalling, summary infographic, and cool-down period.

Experimental design

This study will be conducted as an experimental randomised control trial with an online sample. Respondents will be randomly allocated to one of twelve treatment groups. They will be presented with the same privacy policy, which is followed by an intervention. The intervention will be based on the group they are assigned. They will then play some lab games and answer a series of questions. After this, they will complete the manipulation checks and finish with the experiment.

Design details

Randomization method
The randomisation will be done by Qualtrics, the survey platform employed for this study.
Randomization unit
Individual
Clusters
Not Applicable
Observations
3000
Arms
250-300
Primary outcomes
Willingness to part with sensitive information, Willingness to share information for personalisation
Secondary outcomes
Manipulation checks
Partners
Name: Center for Social and Behavioural Change Partner_type: private_company Url: https://www.csbc.org.in

Cited by

No citation data available for this registration yet.

Keywords