Complexity Preferences over Information Search Methods

How to cite

Shadforth, Christopher; Zizzo, Daniel; Friesen, Lana; Peeters, Luk. 2022. "Complexity Preferences over Information Search Methods." AEA RCT Registry. AEARCTR-0010223. https://doi.org/10.1257/rct.10223-1.0

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

Christopher Shadforth* , Daniel Zizzo , Lana Friesen , Luk Peeters

* Primary investigator

Registry ID
AEARCTR-0010223
DOI
https://doi.org/10.1257/rct.10223-1.0
First registered
Last updated
Status
In development
Trial dates
2022-10-24 – 2023-12-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/10223

Abstract

People have preferences over complexity, as shown in previous studies of complexity preferences over gambles. However, complexity preferences over information methods – ways of receiving information about the state of the world – have not been studied. This is despite the ubiquity of situations where people have a choice between information methods. In this experiment, preferences over complexity of information methods are measured by participants making choices between pairs of information methods that differ in informativeness and/or complexity. After choosing an information method, they then use it in a ball-and-urn belief updating task to estimate the probability that a particular urn was chosen. Differences in choices between pairs of information methods allow observation of complexity preferences.

Experimental design

This economic lab experiment measures complexity preferences between different information methods. Information methods are any way of receiving information about the world. In this experiment, the information methods are sets of two urns mapped to different states of the world, and filled with different coloured balls. The information methods differ in the informativeness of signals they provide and the complexity of using the information method. Participants have an opportunity to use each information method in a belief updating task in the first part of the experiment. Then in a later part of the experiment, they are presented with pairs of information methods that differ in their informativeness and/or complexity, followed by using the chosen information method in a belief updating task. Participants play six pairs of information methods, repeated over four stages. At other points in the experiment, the participants are also asked about their level of anxiety and confidence in themselves; their mathematical ability; their risk preferences; and general demographic factors. The perceived complexity of the information methods used is also measured.

Design details

Randomization method
Randomisation is done by computer.
Randomization unit
Individual
Clusters
At least 60 (a cluster per participant)
Observations
At least 60 participants providing a minimum of 4 rounds of 6 choices, for a total of at least 1440 observations
Arms
At least 60 participants, all playing the same 6 within treatments (choices between information methods)
Primary outcomes
Choices over information methods that differ on complexity and/or informativeness

Cited by

No citation data available for this registration yet.

Keywords