Pilot on Complexity on Social Learning

How to cite

Nei, Stephen; Vorjohann, Pauline. 2024. "Pilot on Complexity on Social Learning." AEA RCT Registry. AEARCTR-0015014. https://doi.org/10.1257/rct.15014-1.0

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

Stephen Nei* , Pauline Vorjohann

* Primary investigator

Registry ID
AEARCTR-0015014
DOI
https://doi.org/10.1257/rct.15014-1.0
First registered
Last updated
Status
Completed
Trial dates
2024-12-13 – 2024-12-31
Countries
United Kingdom of Great Britain and Northern Ireland ()
Registry record
http://www.socialscienceregistry.org/trials/15014

Abstract

Building on a test of how individuals anticipate and account for selection of others into sharing information, we run a pilot to investigate the role of complexity. By changing the distribution of private signals, we make it "obvious" how individuals with strong signals should act and so want to compare the resulting general behavior in this "simple" setting in comparison to the previous more "complex" setting.

Intervention

In comparison to previous stages of this project, we make the private signals easier to understand

Experimental design

Participants will be shown marbles drawn from one of two bags. They will guess the bag. They will then be given the chance to pay to use the first guess of another participant (either drawn randomly from all other participants, or restricted to those who have likewise paid to be matched, or to be matched to a computer "guess" based on an additional draw from the bag) to update their guess about which bag was chosen for them. Beliefs and estimation of how optimal decisions were will also be collected.

Design details

Randomization method
Randomization done by computer selecting which draws the participant sees, so that assignment to signal strength is determined by an individual random draw
Randomization unit
individual for assignment to signal strength, by blocks to the order of selection versus no selection treatment (one ordering is used until half the desired subjects have participated, then the other ordering is used for the remaining subjects)
Clusters
Roudhly 100 Prolific subjects (final amount determined by funding)
Observations
roughly 100 Prolific users
Arms
roughly 50 participants play 2 rounds of the S treatment, then 2 rounds of the N treatment, then 2 rounds of the C treatment. roughly 50 participants play 2 rounds of the N treatment, then 2 rounds of the S treatment, then 2 rounds of the C treatment
Primary outcomes
The difference between the "selection" treatment and the "no-selection" treatment in participants' willingness to pay for the social signal and in how often they comply and update their beliefs conditional on the social signal, and how this difference compares to the difference found when using a more complex bag in previous rounds of the project.

Cited by

No citation data available for this registration yet.

Keywords