Temptation and self-control in intrinsic preferences for information

How to cite

Roy-Chowdhury, Vivek. 2023. "Temptation and self-control in intrinsic preferences for information." AEA RCT Registry. AEARCTR-0010865. https://doi.org/10.1257/rct.10865-2.1

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

Vivek Roy-Chowdhury*

* Primary investigator

Registry ID
AEARCTR-0010865
DOI
https://doi.org/10.1257/rct.10865-2.1
First registered
Last updated
Status
Completed
Trial dates
2023-02-01 – 2023-02-03
Countries
United States of America ()
JEL codes
D83
Registry record
http://www.socialscienceregistry.org/trials/10865

Abstract

By bringing a test of temptation and self-control preferences to the domain of information acquisition, I collect experimental evidence on whether individuals experience self-control problems relating to information.

Experimental design

The experiment occurs over two sessions, conducted online.
In session 1, participants make a choice with a small probability of a negative consequence if a selfish action is taken. Afterwards, they discover that in session 2 they will be offered information on whether the consequence did occur. They are then offered commitment devices to seeing or not seeing the information.
In session 2, depending on commitments made in session 1, information is offered to participants and they choose whether to see it or not.

Design details

Randomization method
Randomisation of treatment groups and prices of information is done in javascript (via Qualtrics) during session 1 itself. Randomisation of implemented preferences, preference payments and information content is done in R using session 1 data.
Randomization unit
Individual
Clusters
600 individuals
Observations
600 individuals
Arms
50:50 split between the two main treatment arms ("info" and "signal") for those participants taking the bonus.
Primary outcomes
The shares of individuals strictly preferring commitments to avoiding or seeing information rather than choosing in the next session (& comparison across treatment arms). Note that there are three possible measurements of preferences, increasing in conservatism: the first is after the basic survey elicitation, the second after a time cost must be borne, and the final accounting for WTA. Each imposes a harsher requirement for preferences to be considered strict. We will conduct analysis across all measures, noting that measure 2 (post time-cost) is our primary focus.
Secondary outcomes
Share of individuals demonstrating dynamic inconsistency in information preferences (& comparison across treatment arms) Final consumption of information across treatments Comparison of preferences across qualitative measures of risk attitudes, time preference, impulse control, altruism, and sentiment about bonus choice.

Cited by

No citation data available for this registration yet.

Keywords