Using the Blotto game to measure beliefs

Last registered on November 17, 2022

View Trial History

Pre-Trial

Trial Information

General Information

Title

Using the Blotto game to measure beliefs

RCT ID

AEARCTR-0010405

Initial registration date

November 13, 2022

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

November 17, 2022, 3:46 PM EST

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country

New Zealand

Region

Primary Investigator

Name

Ronald Peeters

Affiliation

University of Otago

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Leonard Wolk

PI Affiliation

Vrije Universiteit Amsterdam

Additional Trial Information

Status

In development

Start date

2022-11-16

End date

2023-05-01

Keywords

Other

Additional Keywords

Blotto, beliefs, wisdom of the few

JEL code(s)

C83, C91, D80

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

Peeters and Wolk (2019) developed and validated Blotto games to elicit expectations. In their experiments they apply their games to elicit expectations about termination times of a time series process. Given that in many studies expectations are elicited on beliefs and norms about behavior, in this study we apply the most successful Blotto game of Peeters and Wolk (2019) to expectations about behaviour, and do this in the context of the Ultimatum Game. To be precise, we apply the Blotto game to measure beliefs about behaviour and beliefs about beliefs about behavior in this game. The former applies to measuring beliefs about verifiable events, the latter about unverifiable events. In the analysis we will particularly focus on the quality of the measured beliefs in the context of the “wisdom of the few” rather than the “wisdom of the crowd”.

External Link(s)

Registration Citation

Citation

Wolk, Leonard and Ronald Peeters. 2022. "Using the Blotto game to measure beliefs." AEA RCT Registry. November 17. https://doi.org/10.1257/rct.10405-1.0

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

Intervention Start Date

2022-11-16

Intervention End Date

2022-12-22

Primary Outcomes

Primary Outcomes (end points)

Finding out in how far the Blotto game is a valid tool to measure beliefs about behavior and beliefs about beliefs about behavior.

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

Finding out the number of individuals needed to measure beliefs accurately.

Secondary Outcomes (explanation)

Experimental Design

The experiment will be conducted online using Qualtrics and participants will be recruited using Prolific. Participants are participating in one of three treatments. In Treatment A, they will be playing a discretized version of the Ultimatum Game as in Trautmann and van de Kuilen (2015). Participants will be playing the game either in the role of Proposer or in the role of Responder. Responders make their decision according to a strategy method. In Treatment B, participants play (in pairs) the Blotto game, where the outcome of the game is in part determined by the outcome of Treatment A. Participants play this game either in the context of the Proposers’ decisions or in the context of the Responders’ decisions. Treatment C is comparable to Treatment B, but instead of the outcome of Treatment A being in part relevant for the outcome of the Blotto game, average behavior within the session will be relevant.

Experimental Design Details

Randomization Method

All randomizations are computerized.

Randomization Unit

Participants are assigned to either the Proposer role or the Responder role. This randomization is on the individual level. In Treatment A, Proposers are paired with Responders (as they play the Ultimatum Game). In Treatments B and C, pairs are formed within Proposers and Responders (as they play the Blotto game). In Treatments B and C participant earnings are subject to chance, for which randomizations are done on pair-level.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

One per decision situation (two for each treatment)

Sample size: planned number of observations

600

Sample size (or number of clusters) by treatment arms

200 (100 for each role)

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

The sample sizes are chosen in order to conveniently identify an effect size down to 0.5 standard deviations on 5% significance with 80% power in a two-sided non-parametric test.

Supporting Documents and Materials

IRB