EconBase
← All papers

Multivariate Tie-breaker Designs

Tim P. Morrison, Art B. Owen

arXiv 21 Feb 2022 · Statistics — Methodology · publishedElectronic Journal of Statistics (2024) · 1 citations (OpenAlex)

arXiv:2202.10030 · PDF · DOI · OpenAlex · Extracted main text

Abstract

In a tie-breaker design (TBD), subjects with high values of a running variable are given some (usually desirable) treatment, subjects with low values are not, and subjects in the middle are randomized. TBDs are intermediate between regression discontinuity designs (RDDs) and randomized controlled trials (RCTs). TBDs allow a tradeoff between the resource allocation efficiency of an RDD and the statistical efficiency of an RCT. We study a model where the expected response is one multivariate regression for treated subjects and another for control subjects. We propose a prospective D-optimality, analogous to Bayesian optimal design, to understand design tradeoffs without reference to a specific data set. For given covariates, we show how to use convex optimization to choose treatment probabilities that optimize this criterion. We can incorporate a variety of constraints motivated by economic and ethical considerations. In our model, D-optimality for the treatment effect coincides with D-optimality for the whole regression, and, without constraints, an RCT is globally optimal. We show that a monotonicity constraint favoring more deserving subjects induces sparsity in the number of distinct treatment probabilities. We apply the convex optimization solution to a semi-synthetic example involving triage data from the MIMIC-IV-ED database.

Citation extraction

44
references
65
in-text mentions
44
distinct cited
4
self-citations
9,790
main-text words

appendix boundary found by appendix_command · 78% of the source is main text. Read the extracted text to check this.

Most heavily cited references

The works this paper leans on most, across its whole bibliography — not restricted to papers in our corpus. Ranked by composite intensity, which combines how often a work is mentioned, how many sections mention it, and how much of that falls in the main text rather than the appendix.

ReferenceIntensityMentionsSectionsMain text
1A. B. Owen and H. Varian (2020) Optimizing the tie-breaker regression discontinuity design self1.00053100%
2H. Li and A. B. Owen (2023) A general characterization of optimality in tie-breaker designs0.84333100%
3R. D. Cook and L. Thibodeau (1980) Marginally restricted d-optimal designs0.73732100%
4C. J. Nachtsheim (1989) On the design of experiments in the presence of fixed covariates0.73732100%
5A. Atkinson, A. Donev, and R. Tobias (2007) Optimum experimental designs, with SAS, volume 34 of Oxford Statistical Science Series0.69351100%
6D. Cook and V. Fedorov (1995) Constrained optimization of experimental design0.64422100%
7A. Goldberger (1972) Selection bias in evaluating treatment effects: Some formal illustrations0.64422100%
8D. M. Kluger and A. B. Owen (2023) Kernel regression analysis of tie-breaker designs self0.64422100%
9J. Yang, A. Mandal, and D. Majumdar (2016) Optimal designs for $2^k$ factorial experiments with binary response0.51121100%
10S. Boyd and L. Vandenberghe (2004) Convex Optimization0.51121100%

Showing the top 10 of 44 scored citations.