EconBase
← All papers

An Automatic Finite-Sample Robustness Metric: When Can Dropping a Little Data Make a Big Difference?

Tamara Broderick, Ryan Giordano, Rachael Meager

arXiv 30 Nov 2020 · Statistics — Methodology · 15 citations (OpenAlex)

arXiv:2011.14999 · PDF · DOI · OpenAlex · Extracted main text

Abstract

Study samples often differ from the target populations of inference and policy decisions in non-random ways. Researchers typically believe that such departures from random sampling -- due to changes in the population over time and space, or difficulties in sampling truly randomly -- are small, and their corresponding impact on the inference should be small as well. We might therefore be concerned if the conclusions of our studies are excessively sensitive to a very small proportion of our sample data. We propose a method to assess the sensitivity of applied econometric conclusions to the removal of a small fraction of the sample. Manually checking the influence of all possible small subsets is computationally infeasible, so we use an approximation to find the most influential subset. Our metric, the "Approximate Maximum Influence Perturbation," is based on the classical influence function, and is automatically computable for common methods including (but not limited to) OLS, IV, MLE, GMM, and variational Bayes. We provide finite-sample error bounds on approximation performance. At minimal extra cost, we provide an exact finite-sample lower bound on sensitivity. We find that sensitivity is driven by a signal-to-noise ratio in the inference problem, is not reflected in standard errors, does not disappear asymptotically, and is not due to misspecification. While some empirical applications are robust, results of several influential economics papers can be overturned by removing less than 1% of the sample.

Citation extraction

73
references
138
in-text mentions
73
distinct cited
6
self-citations
21,878
main-text words

appendix boundary found by appendix_command · 87% 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
1Hampel, F., Ronchetti, E., Rousseeuw, P., and Stahel, W (1986) Robust statistics: the approach based on influence functions1.00093100%
2Angelucci, M., Karlan, D., and Zinman, J (2015) Microcredit impacts: Evidence from a randomized microcredit program placement experiment by Compartamos Banco0.92843100%
3Finkelstein, A., Taubman, S., Wright, B., Bernstein, M., Gruber, J.,… (2012) The Oregon health insurance experiment: Evidence from the first year0.87482100%
4Angelucci, M. and De Giorgi, G (2009) Indirect effects of an aid program: How do cash transfers affect ineligibles' consumption?0.87462100%
5Meager, R (2019) Understanding the average impact of microcredit expansions: A Bayesian hierarchical analysis of seven randomized experiments self0.87462100%
6Meager, R (2020) Aggregating distributional treatment effects: A Bayesian hierarchical analysis of the microcredit literature self0.87452100%
7Belsley, D., Kuh, E., and Welsch, R (1980) Regression diagnostics: Identifying influential data and sources of collinearity0.81142100%
8Hampel, F (1974) The influence curve and its role in robust estimation0.73732100%
9Krantz, S. and Parks, H (2012) The implicit function theorem: History, theory, and applications0.73732100%
10Young, A (2019) Consistency without inference: Instrumental variables in practical application0.73732100%

Showing the top 10 of 73 scored citations.

Cited by, within the corpus

arXiv econ.EM papers that cite this one, ranked by how heavily they lean on it.

Citing paperIntensityMentionsSections
1Sample Fit Reliability1.00054
2Testing Most Influential Sets1.00053
3Finding Most Influential Sets0.87464
4Generalizability with ignorance in mind: learning what we do (not) know for archetypes discovery0.64422
5How Robust are Robustness Checks?0.40511