Motivated vs. Skeptical Beliefs

How to cite

Hagenbach, Jeanne; Saucet, Charlotte. 2021. "Motivated vs. Skeptical Beliefs." AEA RCT Registry. AEARCTR-0007541. https://doi.org/10.1257/rct.7541-1.2

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

Jeanne Hagenbach* , Charlotte Saucet

* Primary investigator

Registry ID
AEARCTR-0007541
DOI
https://doi.org/10.1257/rct.7541-1.2
First registered
Last updated
Status
Completed
Trial dates
2021-03-29 – 2022-05-31
Countries
Germany (Berlin)
Registry record
http://www.socialscienceregistry.org/trials/7541

Abstract

This study investigates a new, psychological, explanation for the fact that the unraveling of information fails in many markets contrary to what standard theory predicts (Milgrom, 1981). In particular, we ask whether it may be that agents are able to interpret skeptically the lack of evidence, but may not want to do so when the unraveling process leads to undesirable truth. To study whether having preferences over belief can explain failure of unraveling, we use a laboratory experiment in the context of a Sender-Receiver game. In a "neutral treatment", we will see if the Receiver is able to undo the Sender's communication strategy to discover the truth as predicted by economic theory. In a "loaded treatment", we will see if this reasoning is made harder by the fact that the information disclosed is about the Receiver's own performance in an IQ-test. We also manipulate exogenously how the lack of evidence should be theoretically interpreted by the Receiver or, said differently, whether being skeptical is psychologically detrimental or beneficial for the Receiver. The resulting dataset will allow us to investigate whether agents are be cognitively able but psychologically reluctant to interpret skeptically the lack of evidence.

Experimental design

We consider one of the simplest incomplete information games, namely a sender-receiver disclosure game: an informed Sender transmits hard information to an uninformed Receiver, who acts upon this information. We exogenously vary wether the information that the Sender communicates about is neutral or loaded, in that it corresponds either to a number with no particular meaning or to the Receiver's performance. In the latter case, the Receiver may have intrinsic preferences for believing he performed well, which potentially affects the way he reads strategically-disclosed information. By playing on the Sender's incentives, we also vary whether vague or absent disclosure should, in theory, be interpreted by the Receiver as a high or low performance.

Design details

Randomization method
We randomly assign each experimental session to one of the 4 treatments. Subjects voluntary sign up to participate in our online sessions but, at the time of registration, cannot know the treatment they will receive.
Randomization unit
Individual.
Clusters
No clustering.
Observations
320 subjects.
Arms
80 subjects per treatment.
Primary outcomes
The key outcome variables are the Senders' information and actions, and Receivers' decisions in each game.

Materials

Cited by

No citation data available for this registration yet.

Keywords