The American Economic Association's registry for randomized controlled trials
NEW UPDATE: Completed trials may now upload and register supplementary documents (e.g. null results reports, populated pre-analysis plans, or post-trial results reports) in the Post Trial section under Reports, Papers, & Other Materials.
Estimating Belief-based Utility
Last registered on July 27, 2020
View Trial History
Estimating Belief-based Utility
Initial registration date
July 27, 2020
July 27, 2020 10:26 AM EDT
Contact Primary Investigator
Other Primary Investigator(s)
Additional Trial Information
We collect data from a laboratory experiment to investigate the belief-based component of the utility function. The experiment consists of two parts: an IQ test and a learning exercise. In the main treatment, every participant observes a signal about his relative performance in the IQ test. The signal is either perfectly informative or entirely uninformative with equal probability. Then the participant is asked to report his subjective probability of the signal being indicative of his performance. In the control treatment, participants do not observe signals but are asked to submit reports for each possible signal realization. We assume that a signal about cognitive ability received in the main treatment affects subjects' belief-based utility, while a signal that is being considered in the control treatment does not. This allows us to empirically investigate the belief-based component of the utility function and estimate its parameters at the population level, as well as for groups of individuals with similar characteristics. Moreover, we plan to use the data from the main treatment to reconcile contradictory evidence on asymmetric updating (Eil and Rao, 2011; Mobius et al, 2014; Schwardmann and van der Weele, 2016; Buser et al., 2018; Coutts, 2018) and further investigate the role of expectations in beliefs formation. In particular, we plan to examine beliefs updating about events to which subjects assign zero prior probability. Last but not least, we use additional survey data to examine the occurrence of time inconsistency in belief formation arising due to anticipatory feelings over upcoming signal realizations. Therefore, we plan to test the psychological expected utility theory formulated by Caplin and Leahy (2001).
Kozakiewicz, Marta. 2020. "Estimating Belief-based Utility." AEA RCT Registry. July 27.
Sponsors & Partners
Intervention Start Date
Intervention End Date
Primary Outcomes (end points)
There are two main outcome variables: 1. the subjective beliefs about one's IQ test result measured before observing a signal, and 2. the subjective probability of the signal being indicative of one's relative performance (measured in the main treatment after observing the signal, and after considering the signal in the control treatment).
Primary Outcomes (explanation)
The final outcome variable will be constructed based on deviations of subjective probability (2) from the Bayesian benchmark based on the subjective beliefs about one's IQ test result (1). The final outcome variable will be compared between subjects in the two treatments.
Secondary Outcomes (end points)
Additionally, we collect data on subjects' personality traits, anxiety (state and trait), test-related emotions experienced during the task and habitual use of emotion regulation strategies.
Secondary Outcomes (explanation)
Participants are randomly assigned to one of two treatments. In the main treatment, every participant observes a signal about his relative performance in the IQ test. The signal is either perfectly informative or entirely uninformative with equal probability. Then the participant is asked to report his subjective probability of the signal being his performance. In the control treatment, participants do not observe signals but are asked to submit reports for each possible signal realization.
Experimental Design Details
randomization done in office by a computer
randomization at the level of experimental session (every session is randomly assigned to be either the treatment or the control); randomization at the individual level in the main treatment (a participant obtains a signal that is equal to his relative performance with probability 0.5 or, with the same probability, equal to a randomly drawn number)
Was the treatment clustered?
Sample size: planned number of clusters
Sample size: planned number of observations
Sample size (or number of clusters) by treatment arms
294 in the main treatment, 102 in the control treatment
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Supporting Documents and Materials
INSTITUTIONAL REVIEW BOARDS (IRBs)
German Association for Experimental Economic Research e.V.
IRB Approval Date
IRB Approval Number
Post Trial Information
Is the intervention completed?
Is data collection complete?
Is public data available?
Reports, Papers & Other Materials
REPORTS & OTHER MATERIALS