Designing Persuasive Experiments

2026-05-15

AI Summary

This study addresses the inherent conflict between experimenters, who seek regulatory approval, and regulators, who aim to maximize social welfare. The authors propose a novel framework in which the regulator sets a minimum expected welfare threshold, and the experimenter optimizes the trial design subject to this constraint without revealing private preferences or costs. This approach uniquely embeds social welfare considerations directly into the experimental design optimization, effectively mitigating strategic Bayesian persuasion concerns. Theoretically, under a normal prior, Neyman allocation is shown to be uniformly optimal. Numerical experiments based on historical clinical trial data demonstrate that, compared to conventional designs, the proposed method reduces average sample sizes by over 48% while maintaining equivalent levels of social welfare.

Abstract

Incentives in experimental design are often misaligned: experimenters design and finance experiments to seek regulatory approval, while regulators seek to maximize social-welfare. We propose a framework to resolve this conflict, wherein regulators set a minimum expected welfare threshold, and experimenters optimize designs subject to this constraint. It requires no knowledge of experimenters' private preferences or costs and mitigates strategic Bayesian persuasion. Under normal priors, sampling according to the Neyman-allocation is always optimal, independent of the specific objectives. Furthermore, we characterize the optimal stopping-rule. In a numerical study calibrated to historical clinical-trial data, our framework reduces expected sample-sizes by over 48% relative to classical designs that attain the same social-welfare.

Problem

• experimental design
• incentive misalignment
• regulatory approval
• social welfare
• Bayesian persuasion

Innovation

• experimental design
• Bayesian persuasion
• Neyman allocation
• optimal stopping rule
• social welfare