Bipartite causal inference with interference, time series data, and a random network

Georgia Papadogeorgou; Zhaoyan Song

arxiv: 2404.04775 · v3 · pith:Q7Y5S3VQnew · submitted 2024-04-07 · 📊 stat.ME

Bipartite causal inference with interference, time series data, and a random network

Zhaoyan Song , Georgia Papadogeorgou This is my paper

Pith reviewed 2026-05-24 02:20 UTC · model grok-4.3

classification 📊 stat.ME

keywords bipartite causal inferenceinterferencetime series dataexposure mappingunconfoundednessrandom networkcarryover effectsmatching estimators

0 comments

The pith

Unconfoundedness of outcome-unit exposures follows from unconfoundedness assumptions on interventional units' treatment assignment and the random network.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper addresses causal inference in bipartite settings where treatments assigned to one set of units affect outcomes for another set through connections on a time-varying network. It defines immediate and carryover causal effects for outcome units as time-averaged contrasts of potential outcomes under different exposure values. The central result shows that exposure unconfoundedness for outcome units can be derived from standard assumptions on treatment assignment to interventional units and on the random network, without needing to assume it directly for exposures. This separation respects the bipartite structure and supports estimation for binary, continuous, and multivariate exposure mappings. For binary cases, the paper gives matching and covariate-balancing algorithms whose bias is bounded.

Core claim

In bipartite causal inference with interference, observational time-series data, and a random network, unconfoundedness of the exposure received by outcome units holds under unconfoundedness assumptions on the interventional units' treatment assignment and the random network. This derivation enables definition of immediate and carryover causal effects for each outcome unit, which are contrasts of potential outcomes under different values of the immediately preceding and past exposures, respectively, averaged over time. The result applies to binary, continuous, and multivariate exposure mappings. In the binary case, algorithms that combine matching and covariate balancing produce estimators,

What carries the argument

Derivation of exposure unconfoundedness for outcome units from assumptions on interventional treatment assignment and random network, within an exposure mapping framework that defines immediate and carryover effects.

If this is right

Immediate and carryover effects can be estimated while respecting the bipartite separation of units.
The unconfoundedness result holds for binary, continuous, and multivariate exposure mappings.
For binary exposure and carryover mappings, matching plus covariate balancing yields estimators whose bias is bounded.
The framework supports analysis of time-series observational data with a changing network.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The separation of assumptions might allow similar unconfoundedness transfers in other partitioned network designs beyond bipartite graphs.
Averaging effects over time could be adapted to settings with irregular observation intervals if the network randomness assumption is maintained.
The bounded-bias estimators could be compared in simulations that vary the degree of network randomness to test sensitivity.

Load-bearing premise

Unconfoundedness assumptions hold for the interventional units' treatment assignment and the random network.

What would settle it

An empirical dataset or simulation in which treatment assignment to interventional units and the network are unconfounded yet the induced exposure for outcome units remains associated with unobserved factors that affect the outcome.

Figures

Figures reproduced from arXiv: 2404.04775 by Georgia Papadogeorgou, Zhaoyan Song.

read the original abstract

In bipartite causal inference with interference, interventional units might receive treatment or control, and they might affect the outcome of outcome units through their connections on a bipartite network. We study bipartite causal inference with interference based on observational data across time and a changing bipartite network. Under an exposure mapping framework, we define the immediate and carryover causal effects for each outcome unit, representing contrasts of potential outcomes under different values of the immediately preceding and past exposures, respectively, averaged over time. We establish unconfoundedness of the exposure received by outcome units based on unconfoundedness assumptions on the interventional units' treatment assignment and the random network, hence respecting the bipartite structure of the problem. Our results hold for binary, continuous, and multivariate exposure mappings. In the special case of binary exposure and carryover mappings, we propose algorithms for the immediate and carryover causal effects that combine matching and covariate balancing. We show that the bias of the resulting estimators is bounded. In our motivating study, we find some evidence that smoke from wildfires has an immediate impact on reducing transportation by bicycle in San Francisco.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper gives a workable extension of exposure mappings to time-series bipartite interference with random networks and proposes matching-balancing estimators with bias bounds, but the unconfoundedness step needs explicit handling of possible network-treatment feedback.

read the letter

The main contribution is a set of definitions for immediate and carryover effects under exposure mappings in a bipartite time-series setting where the network changes each period. They derive unconfoundedness for outcome-unit exposures from standard unconfoundedness on the interventional units plus randomness of the bipartite network, and they give matching-plus-balancing estimators whose bias is bounded for the binary case. The wildfire-smoke application to bicycle counts in San Francisco is a direct illustration of the setup. These pieces together look new relative to the cited literature on static or non-bipartite interference. The framework covers binary, continuous, and multivariate exposures, which broadens its reach. The bias bounds are a practical plus for applied work. The central unconfoundedness claim is stated cleanly and respects the bipartite split, which is the right way to set it up. The stress-test concern about network randomness possibly depending on past treatments is worth checking in the proofs; if the paper only assumes conditional independence at each t without ruling out feedback into the network process, the carryover identification could slip. The empirical claim is modest (“some evidence”), so the application functions more as a proof of concept than a strong demonstration. This is aimed at causal-inference researchers who already work with networks and longitudinal data, especially in environmental or social applications. A reader who needs to adapt exposure mappings to changing bipartite graphs will find usable definitions and estimators here. The paper deserves peer review because the core technical moves are clear enough to be checked and the setting is common enough that referees can judge whether the assumptions are reasonable in practice.

Referee Report

2 major / 2 minor

Summary. The paper develops a framework for bipartite causal inference with interference using time-series observational data and a random, time-varying bipartite network. Under an exposure mapping, it defines immediate and carryover causal effects for outcome units as time-averaged contrasts of potential outcomes. The central result establishes unconfoundedness of outcome-unit exposures from unconfoundedness of interventional-unit treatment assignments plus randomness of the network, respecting the bipartite structure. This holds for binary, continuous, and multivariate exposures. For the binary case, matching-plus-covariate-balancing estimators are proposed with bounded bias. An application to wildfire smoke effects on San Francisco bicycle ridership is presented.

Significance. If the unconfoundedness derivation is valid under the stated assumptions, the work provides a principled extension of causal inference methods to bipartite interference settings with temporal dynamics and carryover, including practical estimators with explicit bias bounds. The allowance for changing networks and multiple exposure types is a strength for applied settings like environmental epidemiology. The bias-bound result for the binary estimators is a concrete, verifiable contribution.

major comments (2)

[unconfoundedness derivation / Theorem on unconfoundedness] The unconfoundedness result (abstract and the derivation referenced in the reader's strongest claim): the claim that outcome-unit exposure is unconfounded rests on unconfoundedness of interventional treatments plus network randomness, but the time-series setting with carryover effects requires an explicit assumption that the network process at time t is independent of the treatment history up to t (conditional on covariates). The skeptic's concern is not addressed in the high-level description; without this, the exposure mapping for carryover effects can remain confounded even under the stated assumptions.
[bias bounds section] Bias bounds for the binary estimators (section on algorithms for immediate and carryover effects): the bounds are described at high level; it is unclear whether they account for estimation error in the network or for post-hoc choices in the matching procedure, which could affect whether the bounds remain valid under the time-varying network.

minor comments (2)

[abstract and methods] The abstract states results hold for binary, continuous, and multivariate exposure mappings, but the algorithms and bias bounds are given only for the binary case; a brief note on why the general case does not receive estimators would improve clarity.
[introduction / definitions] Notation for the exposure mappings and the distinction between immediate and carryover effects should be introduced with an equation or diagram early in the paper to aid readability.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments on our manuscript. We address each major comment below. We will revise the paper to make the relevant assumptions explicit and to clarify the scope of the bias bounds.

read point-by-point responses

Referee: The unconfoundedness result (abstract and the derivation referenced in the reader's strongest claim): the claim that outcome-unit exposure is unconfounded rests on unconfoundedness of interventional treatments plus network randomness, but the time-series setting with carryover effects requires an explicit assumption that the network process at time t is independent of the treatment history up to t (conditional on covariates). The skeptic's concern is not addressed in the high-level description; without this, the exposure mapping for carryover effects can remain confounded even under the stated assumptions.

Authors: We agree that an explicit statement of conditional independence between the network process at t and treatment history up to t is needed to fully address carryover effects in the time-series setting. While our 'random network' assumption is intended to capture this, the high-level presentation does not make the independence explicit. We will add a formal assumption (e.g., Assumption X: network_t ⊥ treatment history | covariates) and revise the unconfoundedness theorem proof to reference it directly. revision: yes
Referee: Bias bounds for the binary estimators (section on algorithms for immediate and carryover effects): the bounds are described at high level; it is unclear whether they account for estimation error in the network or for post-hoc choices in the matching procedure, which could affect whether the bounds remain valid under the time-varying network.

Authors: The bias bounds are derived assuming the network is observed without error and that the matching procedure is fixed ex ante (non-adaptive). They do not incorporate network estimation error. We will revise the section to state these conditions explicitly, note that the bounds apply to the observed time-varying network as given, and clarify that post-hoc data-dependent matching choices are outside the current guarantee. revision: yes

Circularity Check

0 steps flagged

Unconfoundedness derivation is assumption-based and self-contained

full rationale

The paper's central result establishes unconfoundedness for outcome-unit exposures directly from stated assumptions on interventional-unit treatment assignment and network randomness. This is a standard identification argument in causal inference that does not reduce any quantity to a fitted parameter, self-definition, or self-citation chain. No equations or steps in the provided abstract or description equate the target result to its inputs by construction. The time-series and bipartite structure are respected by the explicit assumptions rather than smuggled in via renaming or prior self-work. The derivation is therefore independent of the target estimands.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central claim rests on unconfoundedness assumptions for treatment assignment to interventional units and for the random network; no free parameters or invented entities are described in the abstract.

axioms (2)

domain assumption Unconfoundedness of interventional units' treatment assignment
Invoked to establish unconfoundedness for outcome-unit exposures
domain assumption Randomness of the bipartite network
Invoked to establish unconfoundedness for outcome-unit exposures

pith-pipeline@v0.9.0 · 5720 in / 1259 out tokens · 20684 ms · 2026-05-24T02:20:59.221402+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

29 extracted references · 29 canonical work pages

[1]

Econometric methods for program evaluation

Alberto Abadie and Matias D Cattaneo. Econometric methods for program evaluation. Annual Review of Economics, 10: 0 465--503, 2018

work page 2018
[2]

Large sample properties of matching estimators for average treatment effects

Alberto Abadie and Guido W Imbens. Large sample properties of matching estimators for average treatment effects. Econometrica, 74 0 (1): 0 235--267, 2006

work page 2006
[3]

Network synthetic interventions: A causal framework for panel data under network interference

Anish Agarwal, Sarah H Cen, Devavrat Shah, and Christina Lee Yu. Network synthetic interventions: A causal framework for panel data under network interference. 2023

work page 2023
[4]

Aronow and Cyrus Samii

Peter M. Aronow and Cyrus Samii. Estimating average causal effects under general interference, with application to a social network experiment. Annals of Applied Statistics, 11: 0 1912--1947, 12 2017

work page 1912
[5]

Controlling the false discovery rate: a practical and powerful approach to multiple testing

Yoav Benjamini and Yosef Hochberg. Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal statistical society: series B (Methodological), 57 0 (1): 0 289--300, 1995

work page 1995
[6]

Time series experiments and causal estimands: Exact randomization tests and trading

Iavor Bojinov and Neil Shephard. Time series experiments and causal estimands: Exact randomization tests and trading. Journal of the American Statistical Association, 114: 0 1665--1682, 10 2019

work page 2019
[7]

Cluster randomized designs for one-sided bipartite experiments

Jennifer Brennan, Vahab Mirrokni, and Jean Pouget-Abadie. Cluster randomized designs for one-sided bipartite experiments. Advances in Neural Information Processing Systems, 35: 0 37962--37974, 2022

work page 2022
[8]

Estimation and inference for synthetic control methods with spillover effects

Jianfei Cao and Connor Dowd. Estimation and inference for synthetic control methods with spillover effects. arXiv preprint arXiv:1902.07343, 2019

work page arXiv 1902
[9]

An approach to causal inference over stochastic networks

Duncan A Clark and Mark S Handcock. An approach to causal inference over stochastic networks. arXiv preprint arXiv:2106.14145, 2021

work page arXiv 2021
[10]

The inclusive synthetic control method

Roberta Di Stefano and Giovanni Mellace. The inclusive synthetic control method. Discussion Papers on Business and Economics, University of Southern Denmark, 14, 2020

work page 2020
[11]

Urban bike and pedestrian activity impacts from wildfire smoke events in seattle, wa

Annie Doubleday, Youngjun Choe, Tania M Busch Isaksen, and Nicole A Errett. Urban bike and pedestrian activity impacts from wildfire smoke events in seattle, wa. Journal of Transport & Health, 21: 0 101033, 2021

work page 2021
[12]

Causal inference with bipartite designs

Nick Doudchenko, Minzhengxiong Zhang, Evgeni Drynkin, Edoardo Airoldi, Vahab Mirrokni, and Jean Pouget-Abadie. Causal inference with bipartite designs. 10 2020

work page 2020
[13]

Identification and estimation of treatment and interference effects in observational studies on networks

Laura Forastiere, Edoardo M Airoldi, and Fabrizia Mealli. Identification and estimation of treatment and interference effects in observational studies on networks. Journal of the American Statistical Association, 116 0 (534): 0 901--918, 2021

work page 2021
[14]

Synthetic control group methods in the presence of interference: The direct and spillover effects of light rail on neighborhood retail activity

Giulio Grossi, Patrizia Lattarulo, Marco Mariani, Alessandra Mattei, and O Oner. Synthetic control group methods in the presence of interference: The direct and spillover effects of light rail on neighborhood retail activity. arXiv preprint arXiv:2004.05027, 2020

work page arXiv 2004
[15]

Design and analysis of bipartite experiments under a linear exposure-response model

Christopher Harshaw, Fredrik S \"a vje, David Eisenstat, Vahab Mirrokni, and Jean Pouget-Abadie. Design and analysis of bipartite experiments under a linear exposure-response model. Electronic Journal of Statistics, 17 0 (1): 0 464--518, 2023

work page 2023
[16]

Matching as nonparametric preprocessing for reducing model dependence in parametric causal inference

Daniel E Ho, Kosuke Imai, Gary King, and Elizabeth A Stuart. Matching as nonparametric preprocessing for reducing model dependence in parametric causal inference. Political analysis, 15 0 (3): 0 199--236, 2007

work page 2007
[17]

Causal inference in the social sciences

Guido W Imbens. Causal inference in the social sciences. Annual Review of Statistics and Its Application, 11, 2024

work page 2024
[18]

Enhancing a geographic regression discontinuity design through matching to estimate the effect of ballot initiatives on voter turnout

Luke Keele, Rocío Titiunik, and José R Zubizarreta. Enhancing a geographic regression discontinuity design through matching to estimate the effect of ballot initiatives on voter turnout. Journal of the Royal Statistical Society, Series A, 178: 0 223--239, 2014

work page 2014
[19]

Estimating causal effects in the presence of partial interference using multivariate bayesian structural time series models

Fiammetta Menchetti and Iavor Bojinov. Estimating causal effects in the presence of partial interference using multivariate bayesian structural time series models. Harvard Business School Technology & Operations Mgt. Unit Working Paper, 0 (21-048), 2020

work page 2020
[20]

Variance reduction in bipartite experiments through correlation clustering

Jean Pouget-Abadie, Kevin Aydin, Warren Schudy, Kay Brodersen, and Vahab Mirrokni. Variance reduction in bipartite experiments through correlation clustering. Advances in Neural Information Processing Systems, 32, 2019

work page 2019
[21]

Inclusion of quasi-experimental studies in systematic reviews of health systems research

Peter C Rockers, John-Arne R ttingen, Ian Shemilt, Peter Tugwell, and Till B \"a rnighausen. Inclusion of quasi-experimental studies in systematic reviews of health systems research. Health Policy, 119 0 (4): 0 511--521, 2015

work page 2015
[22]

Constructing a control group using multivariate matched sampling methods that incorporate the propensity score

Paul R Rosenbaum and Donald B Rubin. Constructing a control group using multivariate matched sampling methods that incorporate the propensity score. The American Statistician, 39 0 (1): 0 33--38, 1985

work page 1985
[23]

Causal inference with misspecified exposure mappings: separating definitions and assumptions

Fredrik S \"a vje. Causal inference with misspecified exposure mappings: separating definitions and assumptions. Biometrika, 111 0 (1): 0 1--15, 2024

work page 2024
[24]

Causal health impacts of power plant emission controls under modeled and uncertain physical process interference

Nathan B Wikle and Corwin M Zigler. Causal health impacts of power plant emission controls under modeled and uncertain physical process interference. arXiv preprint arXiv:2306.05665, 2023

work page arXiv 2023
[25]

Bipartite interference and air pollution transport: Estimating health effects of power plant interventions

Corwin Zigler, Vera Liu, Fabrizia Mealli, and Laura Forastiere. Bipartite interference and air pollution transport: Estimating health effects of power plant interventions. 12 2020

work page 2020
[26]

Bipartite causal inference with interference

Corwin M Zigler and Georgia Papadogeorgou. Bipartite causal inference with interference. Statistical science: a review journal of the Institute of Mathematical Statistics, 2021

work page 2021
[27]

Using mixed integer programming for matching in an observational study of kidney failure after surgery

Jos \'e R Zubizarreta. Using mixed integer programming for matching in an observational study of kidney failure after surgery. Journal of the American Statistical Association, 107 0 (500): 0 1360--1371, 2012

work page 2012
[28]

Stronger instruments via integer programming in an observational study of late preterm birth outcomes

Jos \'e R Zubizarreta, Dylan S Small, Neera K Goyal, Scott Lorch, and Paul R Rosenbaum. Stronger instruments via integer programming in an observational study of late preterm birth outcomes. The Annals of Applied Statistics, pages 25--50, 2013

work page 2013
[29]

Zubizarreta

José R. Zubizarreta. Stable weights that balance covariates for estimation with incomplete outcome data. Journal of the American Statistical Association, 110: 0 910--922, 7 2015

work page 2015

[1] [1]

Econometric methods for program evaluation

Alberto Abadie and Matias D Cattaneo. Econometric methods for program evaluation. Annual Review of Economics, 10: 0 465--503, 2018

work page 2018

[2] [2]

Large sample properties of matching estimators for average treatment effects

Alberto Abadie and Guido W Imbens. Large sample properties of matching estimators for average treatment effects. Econometrica, 74 0 (1): 0 235--267, 2006

work page 2006

[3] [3]

Network synthetic interventions: A causal framework for panel data under network interference

Anish Agarwal, Sarah H Cen, Devavrat Shah, and Christina Lee Yu. Network synthetic interventions: A causal framework for panel data under network interference. 2023

work page 2023

[4] [4]

Aronow and Cyrus Samii

Peter M. Aronow and Cyrus Samii. Estimating average causal effects under general interference, with application to a social network experiment. Annals of Applied Statistics, 11: 0 1912--1947, 12 2017

work page 1912

[5] [5]

Controlling the false discovery rate: a practical and powerful approach to multiple testing

Yoav Benjamini and Yosef Hochberg. Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal statistical society: series B (Methodological), 57 0 (1): 0 289--300, 1995

work page 1995

[6] [6]

Time series experiments and causal estimands: Exact randomization tests and trading

Iavor Bojinov and Neil Shephard. Time series experiments and causal estimands: Exact randomization tests and trading. Journal of the American Statistical Association, 114: 0 1665--1682, 10 2019

work page 2019

[7] [7]

Cluster randomized designs for one-sided bipartite experiments

Jennifer Brennan, Vahab Mirrokni, and Jean Pouget-Abadie. Cluster randomized designs for one-sided bipartite experiments. Advances in Neural Information Processing Systems, 35: 0 37962--37974, 2022

work page 2022

[8] [8]

Estimation and inference for synthetic control methods with spillover effects

Jianfei Cao and Connor Dowd. Estimation and inference for synthetic control methods with spillover effects. arXiv preprint arXiv:1902.07343, 2019

work page arXiv 1902

[9] [9]

An approach to causal inference over stochastic networks

Duncan A Clark and Mark S Handcock. An approach to causal inference over stochastic networks. arXiv preprint arXiv:2106.14145, 2021

work page arXiv 2021

[10] [10]

The inclusive synthetic control method

Roberta Di Stefano and Giovanni Mellace. The inclusive synthetic control method. Discussion Papers on Business and Economics, University of Southern Denmark, 14, 2020

work page 2020

[11] [11]

Urban bike and pedestrian activity impacts from wildfire smoke events in seattle, wa

Annie Doubleday, Youngjun Choe, Tania M Busch Isaksen, and Nicole A Errett. Urban bike and pedestrian activity impacts from wildfire smoke events in seattle, wa. Journal of Transport & Health, 21: 0 101033, 2021

work page 2021

[12] [12]

Causal inference with bipartite designs

Nick Doudchenko, Minzhengxiong Zhang, Evgeni Drynkin, Edoardo Airoldi, Vahab Mirrokni, and Jean Pouget-Abadie. Causal inference with bipartite designs. 10 2020

work page 2020

[13] [13]

Identification and estimation of treatment and interference effects in observational studies on networks

Laura Forastiere, Edoardo M Airoldi, and Fabrizia Mealli. Identification and estimation of treatment and interference effects in observational studies on networks. Journal of the American Statistical Association, 116 0 (534): 0 901--918, 2021

work page 2021

[14] [14]

Synthetic control group methods in the presence of interference: The direct and spillover effects of light rail on neighborhood retail activity

Giulio Grossi, Patrizia Lattarulo, Marco Mariani, Alessandra Mattei, and O Oner. Synthetic control group methods in the presence of interference: The direct and spillover effects of light rail on neighborhood retail activity. arXiv preprint arXiv:2004.05027, 2020

work page arXiv 2004

[15] [15]

Design and analysis of bipartite experiments under a linear exposure-response model

Christopher Harshaw, Fredrik S \"a vje, David Eisenstat, Vahab Mirrokni, and Jean Pouget-Abadie. Design and analysis of bipartite experiments under a linear exposure-response model. Electronic Journal of Statistics, 17 0 (1): 0 464--518, 2023

work page 2023

[16] [16]

Matching as nonparametric preprocessing for reducing model dependence in parametric causal inference

Daniel E Ho, Kosuke Imai, Gary King, and Elizabeth A Stuart. Matching as nonparametric preprocessing for reducing model dependence in parametric causal inference. Political analysis, 15 0 (3): 0 199--236, 2007

work page 2007

[17] [17]

Causal inference in the social sciences

Guido W Imbens. Causal inference in the social sciences. Annual Review of Statistics and Its Application, 11, 2024

work page 2024

[18] [18]

Enhancing a geographic regression discontinuity design through matching to estimate the effect of ballot initiatives on voter turnout

Luke Keele, Rocío Titiunik, and José R Zubizarreta. Enhancing a geographic regression discontinuity design through matching to estimate the effect of ballot initiatives on voter turnout. Journal of the Royal Statistical Society, Series A, 178: 0 223--239, 2014

work page 2014

[19] [19]

Estimating causal effects in the presence of partial interference using multivariate bayesian structural time series models

Fiammetta Menchetti and Iavor Bojinov. Estimating causal effects in the presence of partial interference using multivariate bayesian structural time series models. Harvard Business School Technology & Operations Mgt. Unit Working Paper, 0 (21-048), 2020

work page 2020

[20] [20]

Variance reduction in bipartite experiments through correlation clustering

Jean Pouget-Abadie, Kevin Aydin, Warren Schudy, Kay Brodersen, and Vahab Mirrokni. Variance reduction in bipartite experiments through correlation clustering. Advances in Neural Information Processing Systems, 32, 2019

work page 2019

[21] [21]

Inclusion of quasi-experimental studies in systematic reviews of health systems research

Peter C Rockers, John-Arne R ttingen, Ian Shemilt, Peter Tugwell, and Till B \"a rnighausen. Inclusion of quasi-experimental studies in systematic reviews of health systems research. Health Policy, 119 0 (4): 0 511--521, 2015

work page 2015

[22] [22]

Constructing a control group using multivariate matched sampling methods that incorporate the propensity score

Paul R Rosenbaum and Donald B Rubin. Constructing a control group using multivariate matched sampling methods that incorporate the propensity score. The American Statistician, 39 0 (1): 0 33--38, 1985

work page 1985

[23] [23]

Causal inference with misspecified exposure mappings: separating definitions and assumptions

Fredrik S \"a vje. Causal inference with misspecified exposure mappings: separating definitions and assumptions. Biometrika, 111 0 (1): 0 1--15, 2024

work page 2024

[24] [24]

Causal health impacts of power plant emission controls under modeled and uncertain physical process interference

Nathan B Wikle and Corwin M Zigler. Causal health impacts of power plant emission controls under modeled and uncertain physical process interference. arXiv preprint arXiv:2306.05665, 2023

work page arXiv 2023

[25] [25]

Bipartite interference and air pollution transport: Estimating health effects of power plant interventions

Corwin Zigler, Vera Liu, Fabrizia Mealli, and Laura Forastiere. Bipartite interference and air pollution transport: Estimating health effects of power plant interventions. 12 2020

work page 2020

[26] [26]

Bipartite causal inference with interference

Corwin M Zigler and Georgia Papadogeorgou. Bipartite causal inference with interference. Statistical science: a review journal of the Institute of Mathematical Statistics, 2021

work page 2021

[27] [27]

Using mixed integer programming for matching in an observational study of kidney failure after surgery

Jos \'e R Zubizarreta. Using mixed integer programming for matching in an observational study of kidney failure after surgery. Journal of the American Statistical Association, 107 0 (500): 0 1360--1371, 2012

work page 2012

[28] [28]

Stronger instruments via integer programming in an observational study of late preterm birth outcomes

Jos \'e R Zubizarreta, Dylan S Small, Neera K Goyal, Scott Lorch, and Paul R Rosenbaum. Stronger instruments via integer programming in an observational study of late preterm birth outcomes. The Annals of Applied Statistics, pages 25--50, 2013

work page 2013

[29] [29]

Zubizarreta

José R. Zubizarreta. Stable weights that balance covariates for estimation with incomplete outcome data. Journal of the American Statistical Association, 110: 0 910--922, 7 2015

work page 2015