Motivated Information Choice

How to cite

Vautrey, Pierre-Luc; Kwon, Spencer; Murdock, William III. 2019. "Motivated Information Choice." AEA RCT Registry. AEARCTR-0004486. https://doi.org/10.1257/rct.4486-1.0

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

Pierre-Luc Vautrey* , Spencer Kwon , William Murdock III

* Primary investigator

Registry ID
AEARCTR-0004486
DOI
https://doi.org/10.1257/rct.4486-1.0
First registered
Last updated
Status
In development
Trial dates
2019-07-25 – 2019-08-23
Countries
Private
JEL codes
C91, D81, D83, D84, D91
Registry record
http://www.socialscienceregistry.org/trials/4486

Abstract

Previous work has shown that people (1) neglect selection when they learn from a sample (2) avoid information that they expect to be negative or misaligned with their preferences and (3) tend to update asymmetrically and conservatively (compared to the Bayesian benchmark) when receiving information that is ego relevant. Although all these results have been individually observed in isolated studies, our study presents a simple experiment that allows us to investigate how these effects could possibly interact and further compound biased beliefs.

Intervention

No Intervention. See Experimental Design.

Experimental design

This experiment aims to jointly measure an individual's information choice, subsequent belief formation, as well as their metacognition (understanding of their own biases) in a motivated setting. The basic experiment follows a 2 by 2 design, with treatment variations across 2 dimensions:
Ego Relevance: Beliefs elicitations are about the participant's performance (Ego-Relevant) or a random number (Ego-Neutral).
Information Choice: Participants select the bias of the source themselves from two options (Chosen-Bias) or are automatically assigned one of the two biases (Forced-Bias).

Design details

Randomization method
Randomization done by a computer for the deployment on MTurk.
Randomization unit
Individual
Clusters
900 individuals
Observations
900 individuals
Arms
300 Ego-Choice , 300 Ego-Forced, 150 Non-Ego-Choice, 150 Non-Ego-Forced
Primary outcomes
Information choice (Yay VS Nay) ; Final Bias ; Updating Biases
Secondary outcomes
Attribution Biases

Cited by

No citation data available for this registration yet.

Keywords