Back to History

Fields Changed

Registration

Field Before After
Abstract We will conduct a lab experiment to evaluate interventions that leverage blockchain features to promote truth-telling. Specifically, we test how information transmission noise can promote truth-telling. We will conduct a lab experiment to evaluate interventions that leverage blockchain features to promote truth-telling. Specifically, we test how information transmission noise can promote truth-telling. Note on changes: This version adjusts the planned sample size from 100 participants to 172 participants. This update is due to a recalculation of statistical power based on data from a pilot experiment.
Last Published May 07, 2026 10:46 AM May 14, 2026 01:07 AM
Primary Outcomes (End Points) 1. Truthfulness of participants 2. Performance-dependent payoffs participants gain from the experiment 1. The lying probability of participants 2. Performance-dependent payoffs participants gain from the experiment, as a welfare metric
Primary Outcomes (Explanation) Two players share information about the state of the world. They have incentives to misreport, while our treatment intervention attenuates the incentives and promote truth-telling. Two players share information about the state of the world. While welfare is maximized when both parties reveal information truthfully, they have incentives to misreport to maximize personal gains. Our intervention is expected to attenuate lying incentives and promote truth-telling. We expect that, if optimally designed, transmission noise improves welfare.
Experimental Design (Public) The experiment follows a 1-by-2 between-subjects design. Participants have incentives to misreport in the baseline (low transmission noise). High transmission noise is expected to reduce the incentives to lie. By randomizing on whether participants receive the intervention (high transmission noise), we expect that lying is less frequent with the intervention than in the baseline. The experiment follows a 1-by-2 between-subjects design, with a control group and a treatment group, where participants in the treatment group receive the intervention (i.e., noise in information transmission) but those in the control group do not.
Planned Number of Clusters 100 participants. 172 participants.
Planned Number of Observations 100 participants. 172 participants.
Sample size (or number of clusters) by treatment arms There are 50 participants in each of the two treatments from the 1-by-2 design, so the total number of participants is 100. There are 86 participants in each of the two groups from the 1-by-2 design, so the total number of participants is 172.
Power calculation: Minimum Detectable Effect Size for Main Outcomes We target at least an 80% chance to detect the treatment effect at the 0.01 statistical significance level. Based on a preliminary pilot of the experiment, this would require at least 86 participants in each group.
Intervention (Hidden) 1. Transmission noise level: high or low The intervention is noise in information transmission. In the control group, a player's message is revealed to the counterparty as-is; in the treatment group, a player's revealed message is flipped as opposed to the message she/he submitted with some probability.
Secondary Outcomes (End Points) 1. Risk aversion 2. Cognitive ability 3. Social preferences 4. Lying aversion
Secondary Outcomes (Explanation) After the main task that delivers the primary outcomes, the experiment includes four auxiliary tasks designed to elicit participants' risk aversion, cognitive ability, social preferences, and lying aversion, following standard methods in the literature. These auxiliary variables would help us to obtain a deeper understanding of the primary outcomes by connecting them to various strands of the literature.
Back to top