SHARP BOUNDS ON THE VARIANCE IN RANDOMIZED EXPERIMENTS

By P. M. Aronow, Donald P. Green, Donald K. K. Lee. 2014

88 citations recorded by the source corpus.

We propose a consistent estimator of sharp bounds on the variance of the difference-in-means estimator in completely randomized experiments. Generalizing Robins [Stat. Med. 7 (1988) 773-785], our results resolve a well-known identification problem in causal inference posed by Neyman [Statist. Sci. 5 (1990) 465-472. Reprint of the original 1923 paper]. A practical implication of our results is that the upper bound estimator facilitates the asymptotically narrowest conservative Wald-type confidence intervals, with applications in randomized controlled and clinical trials.

Subfields: Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/fe23b1cf82de299fb05a17d7986b869ad1415a5f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Statistics and Probability

Citations in the corpus

Regression‐assisted inference for the average treatment effect in paired experiments (2016, background): “In completely randomized experiments, Aronow et al. (2014) present sharp variance estimators for the difference-in-means estimator of the sample average treatment effect without covariance adjustment.”
Rerandomization and regression adjustment (2019, background): “In general, these confidence intervals can be conservative, and can be improved (Aronow et al. 2014; Fogarty 2018; Ding et al. 2019).”
Bridging Finite and Super Population Causal Inference (2017, background): “Using Freedman (2008a)’s results, Aronow et al. (2014) considered a super population with N units, with the finite population being a simple random sample of size n.”
Design-based Analysis in Difference-In-Differences Settings with Staggered Adoption (2018, background): “, Aronow et al. [2014], but in general those gains are modest.”
Decomposing Treatment Effect Variation (2016, methodology): “In general, we believe this is a natural setup for researchers to formulate and investigate a broad range of questions about impact heterogeneity (e.g., Heckman, Smith, and Clements 1997).”
Inference on a New Class of Sample Average Treatment Effects (2017, methodology): “More efficient CIs can be constructed by directly using sharper bounds on the unobserved correlation between potential outcomes (Aronow, Green, and Lee 2014).”
Covariate-adjusted Fisher randomization tests for the average treatment effect (2020, unknown): “Aronow et al. (2014) proposed an improvement on s̃en under the finite population framework, which can be extended to s̃e∗ (∗ = r, f, l) as well.”
A paradox from randomization-based causal inference (2014, methodology): “Recently, for general outcomes Aronow et al. (2014) derived the sharp bound of S2τ based on the marginal distributions of the treatment and control potential outcomes using the Frechét–Hoeffding bounds (Nelsen, 2007); for binary outcomes Robins (1988) and Ding and Dasgupta (2016) gave simple forms.”
General Forms of Finite Population Central Limit Theorems with Applications to Causal Inference (2016, background): “The finite population asymptotic scheme embeds ΠN into a hypothetical infinite sequence of finite populations with increasing sizes, and the asymptotic distribution of any sample quantity is its limiting distribution along this hypothetical infinite sequence (cf. Lehmann 1999, 1975; Ding 2016; Arono”
Randomization Tests for Weak Null Hypotheses in Randomized Experiments (2018, methodology): “Aronow et al. (2014) brings up tight bounds for covariance estimation in treatment-control randomized experiments with J = 2.”