A Theory of Experimenters: Robustness, Randomization, and Balance

• 成果类型: Article
• 署名作者: Banerjee, Abhijit, V; Chassang, Sylvain; Montero, Sergio; Snowberg, Erik
• 署名单位: Massachusetts Institute of Technology (MIT); New York University; University of Rochester; University of British Columbia
• 刊物名称: AMERICAN ECONOMIC REVIEW
• ISSN/ISSBN: 0002-8282
• DOI: 10.1257/aer.20171634
• 发表日期: 2020
• 页码: 1206-1230
• 关键词: SEQUENTIAL CLINICAL-TRIALS; information acquisition; BIAS; impacts; designs; HEALTH; worms

摘要

This paper studies the problem of experiment design by an ambiguity-averse decision-maker who trades off subjective expected performance against robust performance guarantees. This framework accounts for real-world experimenters' preference for randomization. It also clarifies the circumstances in which randomization is optimal: when the available sample size is large and robustness is an important concern. We apply our model to shed light on the practice of rerandomization, used to improve balance across treatment and control groups. We show that rerandomization creates a trade-off between subjective performance and robust performance guarantees. However, robust performance guarantees diminish very slowly with the number of rerandomizations. This suggests that moderate levels of rerandomization usefully expand the set of acceptable compromises between subjective performance and robustness. Targeting a fixed quantile of balance is safer than targeting an absolute balance objective.