Jason Hartford, Victor Veitch, Dhanya Sridhar, Kevin Leyton-Brown
arXiv 19 Jun 2020 · Statistics — Methodology · 6 citations (OpenAlex)
arXiv:2006.11386 · PDF · DOI · OpenAlex · Extracted main text
Instrumental variable methods provide a powerful approach to estimating causal effects in the presence of unobserved confounding. But a key challenge when applying them is the reliance on untestable "exclusion" assumptions that rule out any relationship between the instrument variable and the response that is not mediated by the treatment. In this paper, we show how to perform consistent IV estimation despite violations of the exclusion assumption. In particular, we show that when one has multiple candidate instruments, only a majority of these candidates---or, more generally, the modal candidate-response relationship---needs to be valid to estimate the causal effect. Our approach uses an estimate of the modal prediction from an ensemble of instrumental variable estimators. The technique is simple to apply and is "black-box" in the sense that it may be used with any instrumental variable estimator as long as the treatment effect is identified for each valid instrument independently. As such, it is compatible with recent machine-learning based estimators that allow for the estimation of conditional average treatment effects (CATE) on complex, high dimensional data. Experimentally, we achieve accurate estimates of conditional average treatment effects using an ensemble of deep network-based estimators, including on a challenging simulated Mendelian Randomization problem.
appendix boundary found by appendix_command · 73% 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 | J. Hartford, G. Lewis, K. Leyton-Brown, and M. Taddy (2017) Deep IV: A flexible approach for counterfactual prediction | 0.888 | 10 | 4 | 70% |
| 2 | W. K. Newey and J. L. Powell (2003) Instrumental variable estimation of nonparametric models | 0.737 | 3 | 2 | 100% |
| 3 | F. P. Hartwig, G. Davey Smith, and J. Bowden (2017) Robust inference in summary data Mendelian randomization via the zero modal pleiotropy assumption | 0.721 | 8 | 3 | 38% |
| 4 | R. Singh, M. Sahani, and A. Gretton (1906) Kernel instrumental variable regression | 0.644 | 2 | 2 | 100% |
| 5 | A. Bennett, N. Kallus, and T. Schnabel (1905) Deep generalized method of moments for instrumental variable analysis | 0.644 | 2 | 2 | 100% |
| 6 | S. Darolles, Y. Fan, J.-P. Florens, and E. Renault (2011) Nonparametric instrumental regression | 0.644 | 2 | 2 | 100% |
| 7 | G. Hemani, J. Bowden, and G. Davey Smith (2018) Evaluating the potential role of pleiotropy in Mendelian randomization studies | 0.644 | 2 | 2 | 100% |
| 8 | G. Lewis and V. Syrgkanis (2018) Adversarial generalized method of moments | 0.644 | 2 | 2 | 100% |
| 9 | T. Dalenius (1965) The Mode–A Neglected Statistical Parameter | 0.511 | 2 | 1 | 100% |
| 10 | J. H. Venter (1967) On Estimation of the Mode | 0.511 | 2 | 1 | 100% |
Showing the top 10 of 29 scored citations.