Hossein Babaei, Sina Alemohammad, Richard Baraniuk
arXiv 25 Oct 2021 · Econometrics · publishedIEEE Transactions on Neural Networks and Learning Systems (2023) · 1 citations (OpenAlex)
arXiv:2110.13262 · PDF · DOI · OpenAlex · Extracted main text
The first step towards investigating the effectiveness of a treatment via a randomized trial is to split the population into control and treatment groups then compare the average response of the treatment group receiving the treatment to the control group receiving the placebo. In order to ensure that the difference between the two groups is caused only by the treatment, it is crucial that the control and the treatment groups have similar statistics. Indeed, the validity and reliability of a trial are determined by the similarity of two groups' statistics. Covariate balancing methods increase the similarity between the distributions of the two groups' covariates. However, often in practice, there are not enough samples to accurately estimate the groups' covariate distributions. In this paper, we empirically show that covariate balancing with the Standardized Means Difference (SMD) covariate balancing measure, as well as Pocock's sequential treatment assignment method, are susceptible to worst-case treatment assignments. Worst-case treatment assignments are those admitted by the covariate balance measure, but result in highest possible ATE estimation errors. We developed an adversarial attack to find adversarial treatment assignment for any given trial. Then, we provide an index to measure how close the given trial is to the worst-case. To this end, we provide an optimization-based algorithm, namely Adversarial Treatment ASsignment in TREatment Effect Trials (ATASTREET), to find the adversarial treatment assignments.
appendix boundary found by appendix_titled_section at “Appendix” · 81% of the source is main text. Read the extracted text to check this.
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.
| Reference | Intensity | Mentions | Sections | Main text | |
|---|---|---|---|---|---|
| 1 | S. Athey and G. Imbens, “The econometrics of randomized experiments,… (2017) The econometrics of randomized experiments | 0.874 | 5 | 2 | 100% |
| 2 | S. J. Pocock and R. Simon, “Sequential treatment assignment with bal… (1975) Sequential treatment assignment with balancing for prognostic factors in the controlled clinical trial | 0.737 | 3 | 2 | 100% |
| 3 | S. J. Pocock, S. E. Assmann, L. E. Enos, and L. E. Kasten, “Subgroup… (2002) Subgroup analysis, covariate adjustment and baseline comparisons in clinical trial reporting: current practiceand problems | 0.693 | 7 | 1 | 100% |
| 4 | E. A. Stuart, “Matching methods for causal inference: A review and a… (2010) Matching methods for causal inference: A review and a look forward | 0.644 | 4 | 1 | 100% |
| 5 | “Enhancing the Outcomes of Low-Birth-Weight, Premature Infants: A Mu… (1990) Enhancing the Outcomes of Low-Birth-Weight, Premature Infants: A Multisite, Randomized Trial | 0.644 | 2 | 2 | 100% |
| 6 | J. L. Hill, “Bayesian nonparametric modeling for causal inference,”… (2011) Bayesian nonparametric modeling for causal inference | 0.644 | 2 | 2 | 100% |
| 7 | D. B. Rubin, “Estimating causal effects of treatments in randomized… (1974) Estimating causal effects of treatments in randomized and nonrandomized studies | 0.644 | 2 | 2 | 100% |
| 8 | U. Shalit, F. D. Johansson, and D. Sontag, “Estimating individual tr… (2017) Estimating individual treatment effect: generalization bounds and algorithms | 0.644 | 2 | 2 | 100% |
| 9 | D. A. Bader, W. E. Hart, and C. A. Phillips, Parallel Algorithm Desi… (2005) Springer, 2005 | 0.511 | 2 | 1 | 100% |
| 10 | J. Clausen, “Branch and bound algorithms-principles and examples,” D… (1999) Branch and bound algorithms-principles and examples | 0.511 | 2 | 1 | 100% |
Showing the top 10 of 67 scored citations.